(avg. read time: 27–53 mins.)
At least the vast majority of Bible versions available today have made people aware that there is controversy surrounding the ending of the Gospel of Mark. These Bibles will still print Mark 16:9–20, and they may even include other ending materials in the main text or in footnotes, but they will have some note about how these verses are absent from key early manuscripts. But if Mark 16:9–20 is not the proper ending written by Mark, how did it get there to then become passed on in the vast majority of manuscripts and printed in (as far as I know) all the modern copies of the Bible and New Testament? If Mark 16:8 is the ending, why would Mark end on such a profoundly unsatisfying note? If neither of these options is viable, does that mean the original ending was lost, and how could that have happened? These are the questions with which many students of the Bible have needed to engage concerning the final chapter of Mark. And they represent the questions we will be engaging in this review of the evidence for the various options to answer our overarching inquiry: how was Mark supposed to end?
My deeper interest in this subject began with my first doctoral seminar that I took while I was a Master of Divinity student at Truett Seminary. I took a seminar on Mark with Dr. Kelly Iverson at the Baylor Department of Religion. We met once a week and one or two of us would be responsible for leading discussions on the subjects for the day in one of the half-sessions (since classes were broken up into two segments with a short break in between). Each of us were required to do two presentations and lead discussions related thereto, which were our secondary graded assignments alongside the primary final paper that we had to write (which turned out to be my first academic journal article to appear in print). The presentation on Mark 16 and the controversy around the ending fell to me. In the interests of meeting doctoral standards for such a presentation, I thought it would be worthwhile to explore the different options for the ending in some depth. The various commentaries we were reading—we were each responsible for representing one scholar alongside our own contributions to discussions—were only divided between the views that Mark 16:8 was the original ending and that the original ending was lost. While I wanted to explore the arguments for those views, I also wanted to try to represent the view that Mark 16:9–20 was, in fact, the original ending. I had never seriously considered the option because I took for granted what so many other scholars had told me about how it was not in the earliest manuscripts and how it must have been a later addition to cope with either an unsatisfying ending in Mark 16:8 or a lost ending. But I thought, even if I was not inclined to agree with the idea, we should at least consider the best arguments for it, as I imagined it was a good possibility that no one else in that class (except for Kelly, perhaps) had seriously considered the position. As I explored those arguments, especially in Nicholas Lunn’s book The Original Ending of Mark,1 I was struck with how solid they were. I was not fully convinced of the account about why it is absent from the major manuscripts it is absent from, but I no longer found tenable the argument against it from internal factors (i.e., that it did not match Mark’s “style” or otherwise fit with Mark’s writing).
That was over seven years ago. Now that I have given the subject further consideration, I cannot say that I have settled all matters related thereto to my own satisfaction, but I am now inclined to agree that Mark 16:9–20 is the genuine ending of Mark. Whether it was part of the original draft or, as David Alan Black argues, it was added at some point later by Mark, I am not as confident about picking between such options, simply because I have not yet weighed and measured the arguments for both views to my satisfaction (or of the view that it was actually written by others based on notes from Mark, so that it still matches his intent to end in this fashion). I am all too aware how this view sets me against many other scholars, including many that I respect and admire deeply, but I think the evidence ultimately weighs in its favor.
To show how I arrived at this conclusion, I do not simply want to present arguments for it, though I do that, of course, in what may be called a “medial” rather than a “maximal” or “minimal” case for the ending of 16:9–20.2 I want to provide a larger review in three broad steps. First, I will lay out the options for the ending. Second, I will outline the arguments for and against the different positions. Third, I will then evaluate the arguments for the different positions. For simplicity, I will not be doing a comprehensive review of the manuscripts, but I will end consideration at an arbitrary but nice round date: 1000 CE. Any manuscript dated after this or for which there is some doubt if it was written before or after the cutoff will not be considered. I am primarily focused on the Greek manuscripts, but I will address others as they come up.
Choose Your Own Ending? Not So Much
According to many scholarly reviews, there are essentially five possible endings, though they are by no means equally possible. First, there is the possibility that Mark ends at v. 8. This is the possibility favored by the majority of scholars today. It is thought that this distinctly unsatisfying ending is what gave rise to the other endings as attempts to make up for it. The likely earliest extant codices and some other scattered manuscripts end Mark at this point.
Second, there is the “shorter ending” of Mark, which is reprinted in some Bibles today. This “shorter ending” can be translated as follows: “But all that they had been told they reported briefly to those who were around Peter. Then after these things, Jesus himself also sent out by means of them, from the east and to the west, the sacred and imperishable [or completely vivifying] proclamation of everlasting salvation. Amen.” There are some variations in the manuscripts that include this text, such as whether or not “Amen” is included, which could seem to be a definitive ending in this context (though this need not be so; see Rom 1:25; 9:5; 11:36; 15:33; Gal 1:5; Eph 3:21; 1 Tim 1:17; 1 Pet 4:11; 5:11; Rev 1:6–7). But despite it often being characterized as an “ending,” it only ends one manuscript: the Old Latin k manuscript. In every other case where it appears, it is included with 16:9–20. Its function appears to be an attempt to be a sort of hinge transition between vv. 8 and 9, which otherwise appears to present a sudden transition.
Third, there is the “longer ending” of Mark, which is the description given to Mark 16:9–20. The vast, vast majority of manuscripts end with this, including almost all of those manuscripts with the shorter ending. We will get into the specifics of this later, but the earliest allusions and quotes we have from Mark 16 among the early church fathers include some reference to Mark 16:9–20. Eusebius and Jerome were aware of controversy around its presence in manuscripts in their own times, but ultimately agreed with majority of witnesses that this ending was authentic to Mark. This properly functions as the ending even in those manuscripts which could be described as having a “merged” ending with the “shorter ending” above.
Fourth, there is a unique addition called the “Freer Logion” between vv. 14 and 15 from Codex Washingtonianus (W). Like with the “shorter ending,” the function of this addition appears to be to break up what is otherwise a sudden transition, this time between Jesus’s rebuke of the disciples and his command to the same. It may be translated as follows:
And they excused themselves, saying, “This age of lawlessness and unbelief is under Satan, who does not allow what lies under the unclean spirit to be prevailed over by the truth and power of God. Therefore, reveal your righteousness now.” Thus were they speaking to Christ. And Christ replied to them, “The term of years of Satan’s power has been fulfilled, but other terrible things are drawing near. And for those who have sinned I was handed over to death, in order that they may return to the truth and sin no more, in order that they may inherit the spiritual and imperishable glory of righteousness that is in heaven.”
As this is a unique addition and has a particular homiletic quality to it, one can see why no one takes it seriously as part of an ending original to Mark.
Fifth, some scholars suggest rejecting all endings and substituting … nothing. That is, some scholars think that Mark’s ending is lost. Whether it is possible to reconstruct the substance of the ending—even if not the details—from parts of Matthew’s and Luke’s endings depends on who one asks. But these scholars do think, based on the hints from Mark’s own Gospel, the structures of the other canonical Gospels, and the implications of early gospel proclamations (such as the likely earliest formally preserved one in 1 Cor 15:3–7) that Mark recorded some appearances.
The Three Basic Positions
As such, with these different options for endings, the five “possible endings” reduce down to three realistic possibilities. Though there are varieties of views I do not thoroughly account for—such as the differences between, e.g., David Alan Black and Nicholas Lunn, despite their agreement that Mark 16:9–20 was by Mark—I break these positions down into three categories. 1) Mark 16:8 is the original, intended ending, and we just need to deal with it. 2) Mark 16:8 is not the original ending, but the original ending is lost. 3) Mark 16:9–20 is the genuine ending of Mark, whether it was part of the original draft, or he added it later (hence why this position is not necessarily about the earliest/original ending if one follows Black’s view).
The case for 1) is built upon a number of arguments. We can set aside the various (and contradictory) accounts for how 16:8 is an adequate ending for Mark, because these accounts are based on the text-critical arguments, rather than serving as justification for the text-critical conclusions.3 That said, the foremost arguments are from external evidence. First, it is often claimed, with figures taken from Bruce Metzger’s A Textual Commentary on the Greek New Testament, that the ending is absent from the two earliest extant Greek codices Sinaiticus (א) and Vaticanus (B), the Old Latin Codex Bobiensis (itk), around one hundred Armenian manuscripts, and the two oldest Georgian manuscripts (dated to 897 and 913).4 One could also add the codex 304 (though it is from after my cutoff date) and one Sahidic manuscript to this. Second, various claims are made as to how late this addition was made, but Metzger simply says, “Clement of Alexandria and Origen show no knowledge of the existence of these verses; furthermore Eusebius and Jerome attest that the passage was absent from almost all Greek copies of Mark known to them.”5 Third, even among those manuscripts that do feature the ending of 16:9–20, it is said that “Not a few manuscripts which contain the passage have scribal notes stating that older Greek copies lack it, and in other witnesses the passage is marked with asterisks or obeli, the conventional signs used by copyists to indicate a spurious addition to a document.”6 Metzger does not list these manuscripts, but Daniel Wallace lists five with markings—138, 264, 1221, 2346, and 2812—and adds a sixth from Parker: 137.7 Fourth, as an appeal to internal evidence, it is often suggested that the vocabulary and style are non-Markan. Fifth, as a further appeal to internal evidence, it is thought that v. 9 simply does not fit with v. 8. As Metzger is the major reference point for most scholars who make this argument, we will once again appeal to his summary:
The connection between ver. 8 and verses 9–20 is so awkward that it is difficult to believe that the evangelist intended the section to be a continuation of the Gospel. Thus, the subject of ver. 8 is the women, whereas Jesus is the presumed subject in ver. 9; in ver. 9 Mary Magdalene is identified even though she has been mentioned only a few lines before (15.47 and 16.1); the other women of verses 1–8 are now forgotten; the use of ἀναστὰς δὲ and the position of πρῶτον are appropriate at the beginning of a comprehensive narrative, but they are ill-suited in a continuation of verses 1–8.8
Sixth, most simply, it is much easier to explain why 16:9–20 would be added than why these verses would be omitted. In fact, Wallace presents this as probably the strongest challenge to those who advocate for the ending of 16:9–20.9
But this position also has several arguments against it. First, as we will see below, the vast, vast majority of manuscript and patristic evidence favors the ending of 16:9–20. The typical—indeed, almost universal—response to this is that “manuscripts must be weighed, not counted.” But this principle is meaningful not in terms of, say, attributing immense significance to one or two manuscripts and little significance to others. It is meaningful in the sense of independent and higher-quality witnesses being preferable to a simply higher quantity of dependent witnesses. And in our case, witnesses of such age (earlier than any manuscript ending at 16:8 and we have no indications among their own manuscripts to the contrary), geographical distribution, and variety (in terms of text-types and text families represented), in addition to sheer number, favor one reading and not another. Second, given what we have seen before about the proclamation of the gospel in 1 Cor 15:3–7 and Acts, it seems peculiar that Mark would find it appropriate to end his Gospel before reporting any appearances of the risen Jesus. Third, given that every other canonical Gospel ends with resurrection appearances, but not the same ones, it also does not make sense that Mark, by most accounts the first of these Gospels written, should be lacking in resurrection appearances when the other writers, for all their differences, thought it best to end something called a “Gospel” with resurrection appearances, as oral gospel proclamations had featured. Fourth, the Gospel according to Mark itself foreshadows such appearances, such reunions of Jesus with his disciples after his death in 9:9–10; 14:28; and 16:7. In fact, 16:8 itself indicates that the initial proclamation of Jesus’s resurrection, as well as the broader one of the disciples in their evangelization, got out somehow. For high-minded modern and postmodern critics who like the provocativeness and ambiguity of an open ending where their imaginations can run wild, this might make sense. For an early Christian author telling the story of the gospel, it does not.10 Fifth, related to this, with the obvious exceptions of Mark 13 and 14:62, there is a noted tendency of Mark to be explicit about the fulfillment of prophecies that fall within his narrative scope. James Snapp, Jr. summarizes the cases well:
Mark 10:33 to 34, for example, is fulfilled in step-by-step detail. The predictive aspect in 11:2 to 3 is fulfilled completely in 11:4 to 6. Jesus’ words in 14:13 to 15 come true in 14:16. After Jesus predicts that “one of the twelve” will betray Him in 14:20, Mark adds, in 14:43, “one of the twelve” when describing Judas Iscariot, even though Judas Iscariot has already been introduced; the reason for the insertion of the phrase is to make explicit the fulfillment of Jesus’ prediction. And, in Mark 14:30, Jesus predicts that Peter will deny Him three times before the rooster crows – a prediction which is fulfilled step-by-step in Mark 14:66 to 72. The reader is thus led to expect an explicit fulfillment of the angel’s prediction that Jesus will be seen in Galilee. With the abrupt ending, however, the expected fulfillment never comes. No stylistic irregularity in Mark 16:9 to 20 is nearly as unMarcan as the irregularity of the abrupt ending.11
Sixth, the ending of Mark 16:8, with the final clause being ἐφοβοῦντο γάρ, is no way to end a Gospel, or, at least in nearly every other case, a proper book.12 Seventh, one must wonder why, when Mark begins on such a positive note, it should end on such a negative note at the moment when there should be exultation in the resolution of the resurrection, particularly for a work designed to be read in Christian assemblies. Eighth, though it is an argument from silence, it is a curious thing that there is no evidence of early interpreters trying to explain the ending at 16:8, with the problems it causes and the lack of satisfaction. Again, modern and postmodern scholars have poured much effort into such explanations, but their ancient predecessors do not appear to have been interested or externally motivated in a similar way. Of course, of all the arguments, this is the one most vulnerable to being overturned with the discovery of other ancient Christian literature, but that seems unlikely, as there are not even secondhand references to such things or responses to the same. Not only do we have no evidence, but the ancient Church seems to have been unaware of it, too.
The case for 2) can be summarized easily enough. One simply needs to combine the arguments for 1) with the arguments against 1). There is no positive textual evidence to establish the argument for the lost ending.
Of course, the inherent weakness of the argument for the lost ending is the fact that there is no positive textual evidence for it. There is no testimony among the Church Fathers to another ending, even vaguely. The supposition that it resembled something like Matt 28 (and possibly something from Luke 24) is just that: a supposition. The further supposition, in which some advocates of this view agree with some advocates of the 16:8 ending, that a scribe composed 16:9–20 on the basis of other Gospels, is ill-justified. The parallels are superficial and anyone who makes such a supposition has difficulty explaining why a scribe with access to any/all of these endings of other Gospels (or Acts) composed 16:9–20 and not something 1) more like any of them, 2) more robust than any of them, or 3) more harmonized with other Gospels than any of them (as we see attempts to harmonize these texts in writing as early as the second century).
I have already stated the case against 3) by virtue of stating the case for 1). The case for 3) is largely built on external evidence, though, as I indicated earlier, it has internal evidence in its favor as well. First, its attestation among the manuscripts is far superior to those attesting to the ending at v. 8. Here is the list of Greek manuscripts (including uncials/majuscules, minuscules, and lectionary texts) that are commonly dated before 1000 that feature 16:9–20 or some portion thereof (owing to damage or a missing page in those cases):
Uncials (28): A (5th), C (5th), D (5th), W (5th), Σ (6th), 099* (6th/7th), 0112* (6th/7th), 0211 (7th), 047 (8th), Ee (8th), Le* (8th), Ψ* (8th/9th), Ge (9th), He (9th), Ke (9th), M (9th), U (9th), V (9th), Y (9th), Δ (9th), Θ (9th), Π (9th), Ω (9th), 0257 (9th), 0287 (9th), X (10th), Γ (10th), S (10th)
Minuscules (77): 33 (9th), 461 (9th), 565 (9th), 566 (9th), 892 (9th), 1143 (9th), 1295 (9th), 2224 (9th), 399 (9th/10th), 1424 (9th/10th), 24 (10th), 27 (10th), 29 (10th), 34 (10th), 27 (10th), 63 (10th), 67 (10th), 92 (10th), 100 (10th), 106 (10th), 135 (10th), 144 (10th), 151 (10th), 161 (10th), 262 (10th), 274* (10th), 299 (10th), 338 (10th), 344 (10th), 364 (10th), 371 (10th), 405 (10th), 411 (10th), 420 (10th), 478 (10th), 481 (10th), 564 (10th), 568 (10th), 584 (10th), 652 (10th), 771 (10th), 773 (10th), 875 (10th), 942 (10th), 1055 (10th), 1076 (10th), 1077 (10th), 1078 (10th), 1079 (10th), 1110^ (10th), 1120 (10th), 1166 (10th), 1172 (10th), 1203 (10th), 1223 (10th), 1225 (10th), 1281 (10th), 1351 (10th), 1392 (10th), 1421 (10th), 1426 (10th), 1452 (10th), 1458 (10th), 1507 (10th), 1582^ (10th), 1816 (10th), 2193 (10th), 2290 (10th), 2324 (10th), 2369 (10th), 2373 (10th), 2414 (10th), 2545 (10th), 2812 (10th), 2860 (10th), 2929 (10th), 2937* (10th)13
Lectionaries (at least 4): l 627 (8th), l 1602 (8th), l 844 (9th), l 2211 (10th)14
The texts above with an * are manuscripts that feature the “shorter ending” in combination with the traditional ending. The texts marked with a ^ are two of the fourteen cases in which there are annotations noting that some manuscripts end at 16:8. I have only featured these two because the other twelve are beyond the timeframe I have designated. But something should be said about these annotations here, for which I quote Snapp’s remarks (I have put in bold the manuscripts listed above):
Manuscript 199, a minuscule from the 1100’s, is related to another manuscript (which is uncial Codex Lambda in Luke and John, and minuscule 566 in Matthew and Luke) which has the Jerusalem Colophon at the end of each Gospel. It has a short note: “In some of the copies this does not occur, but it stops here” (that is, at the end of 16:8).
Manuscripts 20, 215, and 300 have the Jerusalem Colophon, and, at or near Mark 16:9, they share a note which says, “From here to the end forms no part of the text in some of the copies. But in the ancient ones, it all appears intact.” Rather than cast doubt on the passage, this note appears to have been written to affirm its authenticity.
Manuscripts 1, 205, 205abs, 209, and 1582 are representatives of the Caesarean text of Mark. They share a note which says, “Now in some of the copies, the evangelist’s work is finished here, and so does Eusebius Pamphili’s Canon-list. But in many, this also appears.”
Manuscripts 15, 22, 1110, 1192, and 1210 share basically the same note, minus the part about the Eusebian Canons: “In some of the copies, the Gospel is completed here, but in many, this also appears.”15
While some appeal to these comments in rather generalized fashion as casting doubt on the authenticity of this text, all of these comments encourage acceptance of the passage.
In addition to these texts, the vast, vast majority of evidence from the ancient versions favors its inclusion. This includes the Latin (the only version to outnumber the Greek manuscripts), Coptic (Sahidic and Bohairic), Syriac (including Peshitta manuscripts), Ethiopian, and Gothic (as well as, though the proportions are not as clear, Armenian and Georgian). Of course, if we were to include manuscripts from 1000 and beyond, the number of witnesses would increase in favor of this ending as opposed to the ending of 16:8 (with a few exceptions, like some Armenian manuscripts).
Second, even earlier than any of these manuscripts, and earlier than the earliest ones missing 16:9–20, are references or allusions to this text beginning in the second century (the manuscripts for the patristic references are, of course, not as old, but there is no indication among the various manuscript traditions that suggests these allusions and references were not original). Such second-century links include Ep. apost. 10–12 (16:10–11, 14–15), 30 (possibly 16:15); Justin Martyr, 1 Apol. 45 (16:20); Res. 9 (16:14, 19);16 Irenaeus, Haer. 2.32.3–4 (16:14, 17–18); 3.10.5 (16:19); and Tatian’s Diatessaron (53; 55).17 Another second-century source, Papias, does not directly speak to the issue of the ending, but his testimony is nevertheless suggestive of something. As is well known in Gospel studies, Papias attests (as preserved in Eusebius, Hist. eccl. 3.39.15) that Mark conveyed what Peter taught him and that he was careful not to omit anything he heard (cf. Irenaeus, Haer. 3.1.1). Many (though certainly nowhere near all) scholars dispute the veracity of the connection of Mark with Peter, but adjudicating on this matter is ultimately beside the point here. The point is that Papias characterizes the Gospel according to Mark as comporting with Peter’s preaching, which implies that he is aware of Mark containing resurrection appearances, given how important such a testimony was in the NT and to Peter specifically (as noted from 1 Cor 15:3–7 and Acts, which is not even to mention the other Gospels). Already by the latter half of the second century, we have indications of the ending’s influence in multiple forms among writers located in Phrygia (Papias), Rome (Justin Martyr), Gaul (Irenaeus), and Assyria (Tatian), as well as the Epistula Apostolorum, which may have been written in Syria, Asia, or Egypt (interestingly, it should be noted that our named authors, except Papias, each had a definite connection with Rome at some point, which many have suggested is the provenance of Mark, but I do not wish to argue one way or the other on this for now).
Beyond these second-century references, I will simply list the texts which reference Mark 16:9–20, whether by quotation or allusion. There are others beyond these examples, but I will here restrict myself to ones my readers can easily find translations of online:
Tertullian, Scorp. 15 (16:18); Apol. 21 (16:15, 19); Hippolytus of Rome, Apostolic Tradition 32.1 (16:18); Did. apost. 23 (16:15, 20); Vincent of Thibaris, Seventh Council of Carthage (ANF 5:569 [16:15–18]); On Rebaptism 9 (16:14); Jerome, Pelag. 2.15 (16:14); Aphrahat, On Faith 17 (16:16–18); Ambrose of Milan, Spir. 2.145 (16:15), 151 (16:15–18); Fid. 1.86 (16.15); Paen. 1.35 (16:17–18); Ephrem the Syrian, Comm. Diatessaron; Chromatius of Aquileia, Sermons on the Gospel of Matthew Prologue 7 (16:19); Apos. Con. 8.1 (16:17–18); John Chrysostom, Hom. 1 Cor. 38.5 (16:9); Augustine, Cons. [Harmony of the Gospels] 3.69–83 [chs. 24–25; utilizing both Greek and Latin manuscripts]; Sermons on the New Testament 21.16 (16:15); An. orig. 2.23 (16:18); Hom. 1 John 4.2 (16:15); Macarius Magnes, Apocritus 3.16, 24 (16:18); Victor of Antioch, Comm. Mark; Prosper of Aquitaine, The Call of All Nations 2.2–3 (16:15); John Cassian, Inc. 7.20 (16:17); Nestorius (quoted in Cyril of Alexandria, Against Nestorius 2.6 [16:20]); Leo the Great, Ep. 120.2 (16:16); Peter Chrysologus, Serm. 82–83; Patrick of Ireland, Conf. 40 (16:15–16); Letter to the Soldiers of Coroticus 20 (Mark 16:15–16); Isho’dad of Merv, Comm. Mark
Additionally, Eusebius (Gospel Problems and Solutions, “To Marinus” 1) and Jerome (Ep. 120), despite using the same comment regarding the absence of this ending from manuscripts, nevertheless accept it, and Jerome even included it in the Vulgate, which corrected Latin manuscripts at many points by reference to the Greek manuscripts Jerome knew (cf. Ep. 27). On this last point, Snapp provides an important note:
Considering that Jerome explicitly states in his Preface to the Four Gospels that he had tried to avoid creating readings which would startle Latin readers, while simultaneously he had used ancient Greek manuscripts to veto Latin exemplars where they disagreed with the ancient Greek manuscripts, it seems not only reasonable but unavoidable to deduce that (A) Mark 16:9 to 20 was in those early Greek manuscripts which Jerome used, and (B) Jerome expected his Latin readers to be accustomed to the presence of Mark 16:9 to 20.18
My summary of lectionary text evidence above is rather understated, as I was only working with the date of composition of particular lectionary manuscripts and not when the lectionary patterns themselves were developed. I was also sticking with texts I could personally (and easily) verify that were available online. Beyond what I have cited above, Lunn notes the following:
(A) The Synaxaria of the Greek church require Mk. 16:9–20 to be read for Matins on Ascension Day.
(B) This same lection was adopted among the Syrians by the Melchite churches.
(C) According to the Evangelistarium used by the Jacobite Copts, Mk. 16:14–20 was read at the liturgy on Ascension Day.
(D) Mk. 16:9–20 constituted the third of eleven lections which were read successively on Sundays at Matins throughout the year; as well as daily throughout Easter week in both Greek and Syrian churches.
(E) Mk. 16:9–20 was read at Matins for the second Sunday after Easter in both Greek and Syrian churches.
(F) In the Monophysite churches of Syria, Mk. 16:9–20 was read Matins on Easter Tuesday.
(G) Augustine writes that these same verses were read publicly during Easter among the churches in Africa. They are also included in the oldest lectionary of the Roman church.19
Considering how early these lectionaries and liturgies were developed, this variety attests to how widespread the awareness of this ending was in various languages and, thus, in the Greek texts that served as their bases in the early centuries of the Church.
Third, after all of that, advocates for this position also insist that the internal evidence confirms the fittingness of 16:9–20 in Mark. On the one hand, despite the awkwardness of some features of this summary ending, it provides the fitting end for the Gospel that 16:8 does not, particularly as it fits with the pattern of gospel proclamations and of Gospel narratives elsewhere in the NT. The problems identified with ending Mark at 16:8 are solved by 16:9–20 (although there is no explicit reference to Galilee). On the other hand, its supposed “non-Markan” style is overstated without due recognition that similar analyses of similar amounts of text in Mark would, in some cases, yield similar results of peculiarity. But no one questions those instances, since there is no external evidence for their insertion in the Gospel. This illustrates again how the external evidence is really the driving force of the argument against this ending, not the supposedly devastating internal evidence. Furthermore, there are plenty of Markan features in this text with precedent elsewhere in Mark (on which, see below). Of course, what one might expect to be the reflexive retort here would show how critics try to eat their cake and have it too here. The “non-Markan” features are taken as evidence of the non-Markan origin of this text. But the features that can be more readily connected with Mark are also taken as evidence of the non-Markan origin, as the scribe tried to fit this text with Mark. This shows the inherent bias many have against this text’s authenticity, typically because they have been educated otherwise.
Fourth, a variety of proposals have been put forward for why the ending was omitted at all, not only in B and א, but in what were apparently several ancient manuscripts that came from the Alexandrian stream specifically, which we will note below. But despite the intuitive force of the argument that it is easier to explain why the ending would be added rather than omitted, it becomes increasingly difficult to explain this ending’s pervasiveness when you consider the following factors. One, this is not a simple matter of setting Byzantine texts (which are naturally more numerous) against Alexandrian and/or non-Byzantine texts in general, but of one reading represented by a variety of text-types (including mixed lineage cases) and families, including of Alexandrian texts, against a reading represented in a small amount of distinctly Alexandrian texts. It is not only the vast, vast majority reading when we take into account the massive production of Byzantine manuscripts in the second millennium, but even before then among manuscripts and patristic witnesses in many settings. Two, we have second-century evidence for this ending, arguably in the earliest decades of the second century with Papias, and no second-century evidence that is unambiguously against it. Three, despite what is made of the comments of Eusebius and Jerome, this ending proliferated so much that it appears not only in so many manuscripts, but also in the works of so many authors in so many locations across multiple languages, as well as the several traditions of liturgies and lectionaries. Four, as far as we know, no one, not even those more knowledgeable about the manuscripts, mounted an articulated defense of 16:8 as the proper ending, as even Eusebius and Jerome accepted 16:9–20. All of this is not easy to explain from the opposite position; because of the complexity of factors involved, it may be more difficult to explain than the alternative.
Evaluation
Let us start with the external evidence, as this is what the case for ending at 16:8 is built on, yet it is also overwhelmingly in favor of 16:9–20. Conversely, there is no external evidence in favor of a lost ending. It is more difficult to explain the age, distribution, and variety of witnesses to 16:9–20 on the theory that it is not original to Mark than it is to explain the few pieces of evidence against its authenticity. As noted, we have only two Greek manuscripts before 1000 that attest to ending Mark at 16:8. B is curious because it leaves an abnormal amount of blank space before beginning Luke, which is a full blank column and then another near-quarter of one (forty-two lines of one column plus eleven of another), which could accommodate 16:9–20. The other cases of blank columns left before other books in B are more explicable. After 2 Esdras, the format of the columns changes for the poetic literature of the Psalms. After Tobit, a new copyist takes over the work. Finally, Bel and the Dragon marks the end of the OT in B. The abnormal blank space after Mark 16:8 in B indicates two things: 1) the scribe’s exemplar lacked 16:9–20, but 2) the scribe recalled the text, and so left space for it, though it was never actually included.
As for א, the situation is more complicated, but it does seem the that the scribe here actively rejected the ending of 16:9–20. The scribe in this case is a proofreader who composed replacement pages for Mark 14:54b–Luke 1:56a. The proofreader seems to have had trouble in filling out enough text to conclude Mark without leaving an awkward space for marking the beginning of Luke. This forced him to conclude his last line of Mark with only five letters (τογαρ) followed by an excessive arabesque to fill the blank space. Thus, both B and א do provide evidence for manuscripts lacking 16:9–20, but they also provide evidence of scribes showing awareness of the ending, with one leaving space and one actively rejecting it.20 As such, they are not such straightforward witnesses against the ending as often portrayed.
But we also need to take account of the patristic evidence, particularly those comments from Eusebius and Jerome that are often highlighted.21 First, Clement and Origen, both from Alexandria, are often highlighted, including in the Metzger quote above, as key witnesses against 16:9–20. But Clement does not reference ch. 16 at all; in fact, he hardly ever cites any of Mark. Likewise, Origen cites Mark the least of all the Gospels. There may, in fact, be allusions to Mark 16:9–20 in the description of Jesus’s witnesses (particularly Mary Magdalene) in Cels. 2.55 (which is from the much earlier Celsus and could potentially show his awareness of the Markan text), 59–60, but this characterization may have been drawn from John as well. As it stands, though, as far as I can see, Origen only quoted from Mark 16 once, from v. 2, in Hom. Exod. 7.7, so he is hardly a witness for Mark ending at 16:8 either.
Second, Eusebius says (in the text cited previously as “To Marinus” in Gospel Problems and Solutions) that “in almost all copies” (σχεδὸν ἐν ἅπασι τοῖς ἀντιγράφοις) of Mark, the Gospel ends at v. 8, which he further reinforces with saying it is “seldom/occasionally” (σπανίως) followed with the pericope in question.22 But we must remember two things about this comment. One, despite this claim, Eusebius accepted this text and proceeded in his argument on that basis in providing a second solution to a difference between Matthew and Mark. Two, Eusebius is not making some objectively universal claim about Greek manuscripts in his time, much less of manuscripts produced since Mark was completed, as if he had access to them all. He was bishop of Caesarea, a major manuscript center, which may well have been where B and א were produced.23 With Alexandrian texts coming here (which scholars recognize B and א as being), it is perhaps unsurprising that the majority of Mark manuscripts that Eusebius was familiar with did not feature the ending. But that is not how Metzger and others who are less cautious reference it (some even saying that “all” manuscripts at this time lacked the ending, in which case one wonders where it came from and why Eusebius had this discussion, but it is not as if Eusebius’s dialogue is fully played out or properly referenced by scholars who appeal to this point). Rather, the suggestion is that the majority of manuscripts up to and including the ones of this time period did not feature the ending. But this hardly seems to be the case, given what we have observed previously. If we look at writers before and up to a century after Eusebius, we see that the Greek manuscripts with this ending of 16:9–20 were reference points for writers or translations used by writers in/from Phrygia (Papias), Syria Palestine (Didascalia Apostolorum, Jerome, John Chrysostom, Apostolic Constitutions, and Victor of Antioch), Gaul (Irenaeus), Assyria (Tatian), possibly Egypt (Epistula Apostolorum, if it is not from Syria or Asia), Africa Proconsularis (Tertullian, Vincent of Thibaris, Augustine, and maybe the author of On Rebaptism), Rome (Justin Martyr, Hippolytus, and Jerome had manuscripts from here), Persia (Aphrahat), Italy (Ambrose of Milan and Chromatius of Aquileia), Mesopotamia (Ephrem the Syrian), the British Isles (Patrick), and Greece (if Macarius Magnes is Macarius of Magnesia or otherwise from Greece). How likely is it that, supposedly, manuscripts that only accounted for a tiny sliver from somewhere in the second to the early fifth centuries were responsible for such a wide distribution of witnesses in multiple languages?
Third, Jerome’s statement from Ep. 120 (to Hedibia) is typically treated as if it is independent testimony from an ancient scholar known for his text-critical work, which just so happens to resemble the verbiage of Eusebius. Jerome’s letter actually matches Eusebius’s responses to his questions beyond the noted verbiage of how many Greek copies lack the ending (he specifies “Greek” since he was writing in Latin). Such similarities of wording and substance fit with Jerome’s admitted tendency (which others will readily recognize when reading his work in light of earlier authors) of borrowing from other writers in what he dictated to his amanuensis without clearly remembering (and thus marking) which words were his and which were theirs (Jerome, Ep. 112.4; also preserved in Augustine, Ep. 75.4). As such, it would be no surprise that he was paraphrasing Eusebius in response to a broad, but somewhat similar question to the sort that Eusebius addressed from Marinus. But even if we were to disregard these things, appealing to Jerome for the argument that the original ending was 16:8 runs into the same problems we found in Eusebius. Of course, Jerome seems to have had access to more manuscripts than did Eusebius, not least since he consulted manuscripts in Rome prior to his coming to Bethlehem. But if he were to have independently said that most manuscripts lacked this ending, we would still need to explain why this overturns the implicit testimony of the variety of authors noted above. We would also need to explain why he ultimately sided with Eusebius in answering according to the belief that this ending belongs in Mark and, going even further, in including it in his Vulgate, for which he consulted Greek manuscripts to correct Latin ones. Some, like Wallace, might wish to posit that Jerome suffered a failure of nerve, so to speak, and did not want to upset his readers too much, despite his (supposed) personal beliefs.24 Such a characterization of Jerome is difficult to take seriously for anyone with passing familiarity with his work. Particularly relevant is how he marked the stories of Bel and the Dragon and Susanna, additions to Daniel in the LXX, with symbols to indicate that these stories were not originally part of the Hebrew texts. When he makes note of this in the prologue to his Commentary on Daniel, he also notes that Eusebius made appeal to the same fact in argument with Porphyry when he criticized Daniel. Neither Jerome nor Eusebius worried about making these remarks for stories that were popular among Christians (and Jerome himself notes that people were critical of this move of his), yet both of them ultimately proceeded to act in affirmation of the ending of 16:9–20. While Jerome marked these additions to Daniel while still keeping them in his Vulgate, he did no such thing for Mark 16:9–20, which he preserved without comment or need for qualification.
As noted previously, I am not focusing on the versions other than the Greek (and this would strengthen the case for Mark 16:9–20 more than it would weaken it anyway), but some readers may be curious about the fact that so many Armenian manuscripts and two early Georgian ones lack 16:9–20 (we have only singular cases from Old Latin, Syriac, and Sahidic). The extensive evidence from Greek witnesses (not to mention earlier Latin, Syriac, and other witnesses) cannot be overturned on the strength of further flung instances from Armenian and Georgian texts, but some account must be made of this evidence. I cannot do better to summarize the issues here than Snapp:
I have not addressed the evidence from the Armenian and Georgian versions because research into these versions has not yet reached a firm conclusion about the contents of Mark 16 in the earliest stages of these versions; Colwell’s investigation, made in 1937, cannot be considered sufficient, considering how many Armenian manuscripts have been catalogued since that time. But a few generalized remarks here may be better than nothing at all. Bruce Metzger’s statement in A Textual Commentary on the Greek New Testament that “about one hundred Armenian manuscripts, and the two oldest Georgian manuscripts (written A.D. 897 and A.D. 913)” do not contain Mark 16:9-20 is technically true. (Metzger’s information was from the 1930’s. Many more Armenian copies have been catalogued since that time.) But he did not inform his readers that all those Armenian manuscripts are medieval. Nor did he mention that the Georgian version was translated from Armenian; as a result, most readers are likely to get the false impression that the Armenian and Georgian versions represent two independent lines of evidence.
In addition, Dr. Metzger did not reveal that hundreds of other Armenian manuscripts include Mark 16:9-20. One of those copies is Matenadaran-2374 (formerly known as Etchmiadsin-229), a Gospels-manuscript produced in 989 which, according to an annotation, was copied from “authentic and old” exemplars. The covers and illustrations that accompany the main part of Matenadaran-2374 are from the 500’s or 600’s, and if they were taken from the manuscript of which Matenadaran-2374 is a copy (which is likely), this implies a line of descent for this particular Armenian manuscript that goes back to the earliest detectable stages of the transmission of the text of the Gospels in Armenian. In addition, Armenian and Georgian patristic writings (already described) add further confirmation that Mark 16:9-20 was used in Armenia and Georgia long before the earliest existing Armenian and Georgian manuscripts of Mark were produced.25
On the matter of external evidence, that leaves the claim about manuscripts marked with “asterisks and obeli.” We have already seen cases that were marked with annotations, but none of them expressed actual doubt about the text’s authenticity. As for the six examples Wallace cites, literally none of them feature “asterisks and obeli” to mark scribal doubt about the text. Either the marks have a different function than Wallace has indicated, or he has mistaken the marks as asterisks or obeli when they are neither. (For a thorough response on this point, see Snapp’s work here, which provides links and images for the cited manuscripts.)
All in all, the external evidence is decidedly in favor of Mark 16:9–20 and it is not even kind of close. There are contraindications, and they do raise questions, but surely there should be limits to the burden that two early Greek codices can bear, since they are the ultimate foundation of the argument for 16:8 as the ending. We have followed the counsel that manuscripts are to be weighed, not merely counted. But I am not sure how one sees the two manuscripts, plus scattered others that are related to them, as “outweighing” the testimony of earlier, more widespread, more diverse, and good quality witnesses (unless one wishes to stretch so far as to dismiss all of the hundreds of witnesses, at least a significant portion of which are independent, as all somehow sharing the supposed lackluster quality compared to two chief manuscripts).
Since the internal evidence ultimately depends on the external evidence for its forcefulness, we could technically consider this matter resolved in favor of 16:9–20, but it would do to explore the issues. Without the perceived strength of the external witness to fall back on, the position that 16:8 is the proper ending could not survive the many problems created by appeals to internal evidence. The creative suggestions of what Mark’s ending could mean are built on the assumption that 16:8 is the ending, and we must deal with it somehow. If it is meant to be a Gospel in a way somehow fitting the gospel proclamations summarized elsewhere in the NT, it is curious that Mark would consider it so important to preserve the story of the widow’s mite for his audience, but not the appearances of the resurrected Jesus, despite how foundational the latter was for the early gospel proclamations (again, as indicated in 1 Cor 15 and Acts). In fact, with the exception of infancy Gospels (the Protevangelium of James and the Infancy Gospel of Thomas), and Gnostic Gospels focused on teachings (Gospel of Thomas, Gospel of Truth, and Gospel of Judas), every reasonably complete Gospel we have from the patristic era, and some that are fragmentary, feature Jesus’s resurrection appearances and/or words thereafter (besides Matthew, Luke, and John, see the Gospel of Hebrews, Gospel of Peter, Gospel of Philip, Gospel of Mary, and Gospel of Nicodemus; probably also implied in the Gospel of the Nazoreans; cf. Epistula Apostolorum; Apocryphon of James; 2 Apocalypse of James; Letter of Peter to Philip). It would be remarkable indeed for what appears to have been the first Gospel written to feature no such thing, which would be against tendencies of gospel proclamations before and Gospel narratives after it.
As for Mark 16:9–20, scholars often point to its supposedly “non-Markan” words. The standard figure cited is that there are sixteen unique items out of 166 total words (depending on variants, of course). This in itself is demonstrative of nothing, and it only has forcefulness when one avoids comparison with other units of Mark. Consider the following cases from Mark:
1:2–11 (seventeen unique words out of 176)
5:24–34 (nine unique words out of 161)
6:21–29 (twelve unique words out of 172)
9:2–10 (eleven unique words out of 157)
9:42–50 (seventeen unique words out of 147)
10:1–12 (twelve unique words out of 153)
12:1–11 (nineteen unique words out of 162)
13:14–23 (fifteen unique words out of 152)
14:1–9 (twenty unique words out of 159)
14:55–65 (eleven unique words out of 179)
15:21–32 (twelve unique words out of 153)
15:42–16:5 (seventeen [technically eighteen due to a word being repeated] unique words out of 175)26
Beyond the amount of unique words, which does not really tell us anything of significance, this ending is also said to lack characteristic Markan elements, such as the historic present, the word εὐθύς, and the word πάλιν. Furthermore, it is thought peculiar that Mark previously always used a compound form of πορεύομαι, but does not do so here (16:10, 12, 15). One could also point to the unique phrase “to those who had been with him” in 16:10.
We can respond to these in reverse order. In the last case, the phrase is unique, but the prepositional phrase had been used several times in Mark with reference to Jesus and his disciples (3:14; 4:36; 5:18, 37, 40; 14:33; cf. Luke 24:33), but the precise grammatical construction would not have occasion to be used in Mark prior to 14:50 when the disciples flee and thus their being “with him” would have a past-perfect tense, so to speak. As for πορεύομαι, it is notable that no other Gospel, or even Acts, uses only the simple or compounded form of the verb, nor is the overall ratio (regardless of preference) with the inclusion of Mark 16:9–20 out of keeping with these books, as we see a ratio of the simple to compounded at 29:7 for Matthew, 3:16 for Mark, 51:16 for Luke, 16:2 for John, and 37:9 for Acts.27 And it is not as if the simple form is inappropriate in any of the instances in this context (cf. Matt 28:11, 16, 19; Luke 24:13, 28; John 20:17). Finally, the supposedly missing characteristic Markan elements that absolutely, without question, need to be there if this text came from Mark, nevertheless are also absent from other texts in the aforementioned list. The word πάλιν also does not appear in 1:2–11; 5:24–34; 6:21–29; 9:2–10; 9:42–50; 13:14–23; 14:1–9; 15:21–32; 15:42–16:5. The word εὐθύς also does not appear in 9:2–10; 9:42–50; 10:1–12; 12:1–11; 13:14–23; 14:1–9; 15:21–32; 15:42–16:5. And the historic present, the most pervasive of these features, does not appear in 1:2–11; 5:24–34; 6:21–29; 12:1–11; 13:14–23; 14:1–9 (in the last unit of comparison, it should be noted that there are three historic presents in 16:1–8 and none in 15:33–47). Of these texts, eight are missing two elements (1:2–11; 5:24–34; 6:21–29; 9:2–10, 42–50; 12:1–11; 15:21–32; 15:42–16:5), and two are missing all three (13:14–23; 14:1–9). In these various ways, the text may be somewhat irregular, but it is not demonstrably more out of character for Mark than the other noted texts. And as one can see from this survey, such arguments as these are inherently weakened by the mere consideration that sometimes words are not always appropriate or preferred for every context, even according to the individual author’s sensibilities.28
On the other hand, there are characteristic Markan terms in use here. First, πρωΐ (16:9) appears more often in Mark than any other NT text (1:35; 11:20; 13:35; 15:1; 16:2). In fact, if 16:9 is included, Mark accounts for half of the uses of the term in the NT. Second, Mark’s preferred term for proclaiming a message, κηρύσσω(1:4, 7, 14, 38–39, 45; 3:14; 5:20; 6:12; 7:36; 13:10; 14:9), appears here (16:15, 20). Third, the word εὐαγγέλιον, a major term for referring to the gospel that appears twice as often in Mark as in the much longer Matthew (1:1, 14–15; 8:35; 10:29; 13:10; 14:9) also appears here (16:15). Fourth, the use of καλῶς is somewhat less distinct for Mark, but it still appears in his Gospel more than the others (7:6, 9, 37; 12:28, 32), including here (16:18). Fifth, this text includes other rare words for the NT that appear elsewhere in Mark, such as ἀπιστία (6:6; 9:24; 16:14 cf. 16:11, 16), σκληροκαρδία (10:5; 16:14), κτίσις(10:6; 13:9; 16:15), ἄρρωστος (6:5, 13; 16:18), and πανταχοῦ (1:28; 16:20). This can also apply to words like μεταμορφόω in 9:2 being similar to ἐν ἑτέρᾳ μορφῇ in 16:12, as well as φανερόω in 16:12, 14 (as well as 4:22) being similar to φανερός in 3:12; 4:22; and 6:14.
As for how the text fits with the preceding part of ch. 16, it must be admitted that the transition is awkward and the ending does not fit as smoothly with the rest of the chapter as the similar empty tomb stories in the other Gospels with their respective chapters (though some say that John 21 fits awkwardly with John 20, even though there is no textual evidence of it being inauthentic). Still, we also see here some indications of how the arguments from internal evidence are often built on the assumption that the external evidence is decisive against 16:9–20. After all, surely the transition between vv. 8 and 9 is too harsh for Mark to have written it and thought it would make sense to not continue the scene with the women. But if that is so, why is it any less of a problem for the idea that a later scribe added the text to an unsatisfying ending? Are we to suppose that this is an “any port in a storm” mentality, that any ending will do and that the scribe who supposedly produced this text just happened to think that way? In any case, this is certainly not the only awkward transition in the NT (e.g., John 14:31–15:1; Acts 12:25 [in the context of 11:27–30 and 13:1]; Rom 10:17). And one aspect about it that may be more easily explicable—the “reintroduction” of Mary Magdalene—is a result of the fact that, for the first time, the narrative focuses on Mary and not the other women, and so Mark gives more information about her (which also links with a prophecy of Jesus in 16:17).
Otherwise, some of the awkward qualities of the ending are explicable by this section having more the character of a summary than a fully fleshed out narrative. For some reason that we can only speculate about, Mark has kept the substance of the ending that is in line with the other Gospels, but has rushed it, so that we never get an explicit note about the disciples meeting Jesus in Galilee like we do in Matt 28:16 (after 28:7, 10). Nor is there a scene spotlighting Peter, like we might expect from 16:7. This “rushed” character is also seen in the telescoping of the narrative, so that what will be fleshed out in Acts is compressed into the last verse, and it otherwise seems as if all of these things happened immediately afterwards. A further indication that we are dealing with a summary is the unusual density of using some form of ἐκείνος as a substantive pronoun (16:10–11, 13, 20; cf. 4:20; 12:4–5).29
This characterization of 16:9–20 as a summary that nevertheless fits with the rest of Mark makes sense of two observations. First, this ending does not appear to have been written to resolve everything in Mark’s Gospel (particularly in the last chapters), to fit smoothly with 16:1–8, nor to fix the potential harmonization difficulties with other Gospel narratives. That is, if this was invented by a later scribe to address such problems, one would expect the scribe to have been more attentive to these issues, but these issues make sense if the author is simply providing a not-so-tidy summary, for whatever reason. Second, the ending still does fit with the Gospel as a whole, reflecting many characteristics of the first chapter. One, the only references to the Lord (κύριος) that are not in dialogue appear in Mark’s opening quotation in 1:3 and in 16:19–20. Two, the opening concerns Jesus’s predecessor in preaching while the closing concerns Jesus’s successors in preaching. Three, the only references to literal baptism (which are not part of John’s appellation, as in 6:14, 24) are in 1:5, 8–9; and 16:16. Four, Mark brackets his narrative with the Spirit’s descent from heaven (1:10) and Jesus’s ascent into heaven (16:19). Five, the verb phrase κηρύσσω τὸ εὐαγγέλιον appears only in 1:14 and 16:15. Such fittingness seems to demonstrate an overarching vision of linking the beginning and the end, but not so much a short-term vision of forming an ending that dots every i and crosses every t of the last few chapters (which otherwise might have been accomplished with more time).
Alternatively, when we consider the issue of Mark potentially ending with ἐφοβοῦντο γάρ, although there are no appropriate analogies for such a thing, the scholars who advocate this ending want it both ways. On the one hand, it must be so obviously an inadequate way for Mark to end his Gospel that it thereby motivated other Synoptic Gospel writers and a later scribe altering Mark to add resurrection appearances. (It also requires that Mark misread his audience so poorly as to think this ending would have the intended effect on them when even more educated and literate scribes would not find it acceptable.) On the other hand, this ending must have been the original because it is so proper to Mark—in whatever number of ways scholars think that is—to end this way and leave the dissatisfaction that drove the rest of the history of Synoptic Gospel endings. Still, one may wonder why this is the last state of the women besides Mary Magdalene, but we can only speculate. Lunn posits an interesting idea on this point that I offer here as a mere possibility:
What might be happening in 16:8, we propose, is something similar to what many have seen in 14:51–52 of the same Gospel. The brief scene there about the young man who fled naked at the arrest of Jesus has been taken as an autobiographical insertion relating Mark’s own experience. In 16:8, the uniqueness of the particular perspective given in comparison with the other Gospels, may suggest that we are in fact seeing the events in question through Mark’s own eyes. We here advance the idea that the women running back from the tomb in fear and astonishment passed by the house of Mark en route for that of John. The young man’s attention being drawn to the excited women, some of whom he recognized, he naturally inquired into what had happened as they rushed by. Not stopping to speak, either to Mark or anyone else in the street, the event of the women’s flight from the tomb was entered in Mark’s own record as “they said nothing to anyone.” He was evidently not present when the report was eventually made to Peter and John, at the latter’s house, though later he doubtless would have heard of its contents, probably from Peter.30
Whatever the case may be as far as this feature is concerned, the internal evidence is once again more in favor of ending at 16:20 than at 16:8. The last unit is not a completely neat fit with the with the last chapters of the Gospel, but it is by all measures a better fit of an ending than 16:8, which causes many more problems. Indeed, I doubt anyone would think 16:8 is an appropriate ending if they did not feel compelled to do so by the external evidence, which, as we have seen already, favors 16:9–20 as being authentic.
But we must face one more question. If 16:9–20 is the authentic ending of Mark, how do we explain how it could have been lost in some manuscripts? I am not particularly confident in the idea that heretical copyists (or copyists sympathetic to heretical ideas) deliberately omitted the ending, since this somehow left the endings of the other Gospels more or less untouched (to the extent that no one doubts these other endings as authentic).31 In fact, I am not particularly settled on any proposal, not least because other theories I have seen are not mutually exclusive, and there could be multiple factors. I offer those here as possibilities. One, the ending could have been lost in some codex copies. Lunn studied a sample of uncials number 01–045 and of minuscules number 1–800, finding that 11 of the uncials featured defective endings of books while 103 of the 800 minuscules had similarly defective endings.32 It is not inconceivable that this ending could have been on one page of one or more of the most ancient manuscripts that ended up lost. Codex Bobiensis (itk) may also provide support for this, since it is the only manuscript that ends like it does with an ending that is otherwise combined with 16:9–20 in manuscripts where it appears, and even some of the manuscripts that feature part of the ending are themselves defective/damaged copies. Even beyond Lunn’s study, one could add from the uncials 099, 0112, and 0257 as having defective endings of Mark alone, despite them featuring some portion of vv. 9–20 (0287 is fragmentary and, remarkably, the only discernible text from Mark is 16:19). Two, a scribe might have misinterpreted lectionary divisions, which were not always consistently applied, wherein ἀρχή and τέλος are applied to the beginning and end of a lectionary reading. But if the latter fell at the end of v. 8, since we have seen how 16:9–20 is treated as a separate lectionary unit, it could have been misinterpreted as the end for Mark and vv. 9–20 would be left out of the copy produced thereafter. Three, the text might have been excised out of apologetic considerations. We have already seen difficulties that Eusebius and Jerome encountered in harmonizing this text with other Gospels (particularly Matthew), and others, recognizing the awkwardness of some elements of the ending and the difficulties in clarity about fitting it with other Gospels, may have opted for excising it altogether to make harmonization somewhat easier. This theory is not special pleading, as we can see that Eusebius and Jerome had no problem with noting that a text was not original to address objections in apologetic contexts. This tactic was rather more legitimate in the case of Bel and the Dragon and Susanna, as those stories were not part of the original version of Daniel known in Hebrew and Aramaic. But someone else, perhaps with reinforcement from the other factors, may have thought it easier to excise this text for similar reasons when dealing with difficulties if he thought there were doubts about it anyway. I would also direct my readers to other apologists, teachers, and formally educated Christians of our own day who, when faced with objections concerning Mark 16:9–20, fall back on the notion that the ending is not original as a sufficient way of addressing those objections.33
With all of these points considered, I think the case favors that Mark 16:9–20 belongs as the authentic ending of Mark. We have not considered all factors related to this idea, such as whether Mark ultimately added it later or if it was there all along as part of a more continuous process, or how exactly the differences arose (as I have not adjudicated between various theories). But the end result would remain the same in any case. Mark 16:9–20 is properly part of Scripture, without qualifications, without brackets. I did not always treat it as such, and I have waited until now to address my reasons for doing so, as well as my reasons for changing my mind. From now on, I will simply refer to this document to explain my reasons and go on treating this text as the part of Scripture it is. I invite others to do the same.34
Nicholas P. Lunn, The Original Ending of Mark: A New Case for the Authenticity of Mark 16:9-20 (Eugene: Pickwick, 2014). Since then, I have read what I regard as an even stronger and more thorough case represented by James Snapp, Jr., “Authentic: The Case for Mark 16:9-20,” May 2015 (available online at: https://www.academia.edu/12545835/Authentic_The_Case_for_Mark_16_9_20).
By “medial” case, I mean something in between a “maximal” case like those presented by Lunn and Snapp and a “minimal” case that would rely only on the earliest evidence.
For such varied accounts, see Robert H. Stein, “The Ending of Mark,” BBR 18 (2008): 86–88. I think Snapp’s characterization of these accounts is apt and accurate to the point of their basis on the external evidence of the ending being at 16:8: “The commentators who regard Mark 16:8 as an intentional ending are in disarray when it comes to proposing Mark’s intention. Even if they were united, though, this would not improve their view. Whatever reasons can be imagined to not describe Jesus’ post-resurrection appearances are overwhelmed by the very strong reasons Mark had to describe them. Christ’s post-resurrection appearances were, as Paul makes clear in First Corinthians 15, a very important concern, and not the sort of thing that Mark, urged by Christians at Rome to provide a written record of Peter’s recollections about Jesus, would flippantly fail to mention just so that he could deliberately perplex his readers” (emphasis original). Snapp, “Authentic,” 22.
Bruce M. Metzger, A Textual Commentary on the Greek New Testament, corrected ed. (Stuttgart: United Bible Societies, 1975), 122–23.
Ibid., 123.
Ibid.
Daniel B. Wallace, “Mark 16:8 as the Conclusion to the Second Gospel,” in Perspectives on the Ending of Mark: 4 Views, ed. David Alan Black (Nashville: Broadman & Holman Academic, 2008), 26–27.
Metzger, Textual Commentary, 125.
Wallace, “Mark 16:8,” 11–14.
In the words of Stein, “Here it should simply be noted that the intended readers/hearers envisioned by these suggestions are for the most part very unlike the actual readers that Mark had in mind. They appear to be more like highly-educated 20th- and 21st-century existentialists than like 1st-century Christians, the great majority of whom could not read or write.” Stein, Ending of Mark, 88.
Snapp, “Authentic,” 150.
The three examples I have seen posited for ending a book with γάρ are dubious (the most complete such list I have seen is from James R. Edwards, The Gospel According to Mark, PNTC [Grand Rapids: Eerdmans, 2002], 501). Plotinus’s Ennead 32.5 and Musonius Rufus’s Tractatus 12 are both part of collections edited by others, and so it is not clear that these authors deliberately ended books in this fashion. Plato’s Protagoras 328c, the only pre-Markan example I have seen cited, features a sentence ending in this fashion, but that sentence does not end the book, and it is not even close. Edwards confused the claim about it ending a discourse with it ending a book. As Kelly Iverson (himself a supporter of the view that 16:8 is the proper ending) notes, whiles sentences ending with a final γάρ were, relatively, most frequent in such philosophical literature, they are not common in any genre: “The fact is concluding γάρ statements are extremely, extremely rare at all times and in all genres” (emphasis original). Kelly R. Iverson, “A Further Word on Final Gar,” CBQ 68 (2006): 79–94 (here 93).
Other less complete minuscules may contain the ending as well, but I was not able to confirm this at this time.
There are not many representatives in this category because 1) lectionaries by nature cite only select portions of texts and 2) Mark in general was not as widely used in lectionaries as the other Gospels. But it should be noted that there are many lectionary texts that I could not personally verify, whether because their contents were not adequately identified or because they were not available online (as far as I could tell), and which fall outside the timeframe I have established. But, as I note later, I am only working with the date of composition of these texts, not the invention of the lectionaries per se (one can find twenty-first-century printings of lectionaries and liturgies developed in the fourth and fifth centuries, after all).
Snapp, “Authentic,” 6.
Note that this is a disputed work.
On the arrangement of the harmony in the Arabic and Latin manuscripts of the Diatessaron, see ibid., 36–38.
Ibid., 64.
Lunn, Original Ending, 56. For more, see Snapp, “Authentic,” 140–44.
For more on the issues with א, see Snapp, “Authentic,” 71–81.
Some authors also mention Cyprian for his silence in citing Mark 16. But he is simply a non-witness rather than a witness positively against 16:9–20. For more on this, see ibid., 53–54.
The text and translation can be found on pp. 96–97 of the pdf at https://www.roger-pearse.com/weblog/2015/12/12/eusebius-of-caesarea-gospel-problems-and-solutions-now-online-in-english/.
J. K. Elliott, “T. C. Skeat on the Origins and Dating of Codex Vaticanus,” in The Collected Biblical Writings of T. C. Skeat, ed. J. K. Elliott, NovTSup 113 (Leiden: Brill, 2004), 281–94.
Wallace, “Mark 16:8” 23. In his words, “call it antiquity, tradition of timidity, or not wanting to rock the boat too much.”
Snapp, “Authentic,” 19–20.
Figures are derived from Lunn, Original Ending, 121–26.
Ibid., 142–43.
For much more detailed stylistic analysis, in terms of parts of speech, participant reference (agents and patients are represented in the lexical and grammatical forms of the language), collocations (groupings of two or more words or phrases whose joining together is deliberate), and a number of specific syntactic structures, see ibid., 166–208.
For more on this characteristic of Mark’s ending, see Snapp, “Authentic,” 154–58.
Lunn, Original Ending, 328–29.
Ibid., 343–52.
Ibid., 338, 352–55.
For more on these factors and others, see Snapp, “Authentic,” 165–81.
For more resources available online, I direct my readers to https://www.bterry.com/articles/mkendsty.htm and http://www.curtisvillecc.com/AuthSupplx.html.