Mark 16:9-20 - The Origin of the Long Ending

Where does Mark end?

Although in most bibles chapter 16 of the Gospel of Mark contains twenty verses, in many Mk 16:9-20 are placed in brackets (or otherwise marked). A few note; “Some of the earliest manuscripts do not include 16:9–20,” and some, e.g. the GNT, suggest an alternative ending. As described in Wikipedia, the text of the end of Mark’s gospel is disputed:

The two oldest manuscripts of Mark 16 (from the 300s) then conclude with verse 8, which ends with the women fleeing from the empty tomb, and saying "nothing to anyone, because they were too frightened." ... Many scholars take Mark 16:8 as the original ending and believe that the longer ending (16:9-20) was a later addition. In this 12-verse passage, the author refers to Jesus' appearances to Mary Magdalene, two disciples, and then the Eleven (the Twelve Apostles excluding Judas). The text concludes with the Great Commission, declaring that believers that have been baptized will be saved while nonbelievers will be condemned, and pictures Jesus taken to Heaven and sitting at the Right Hand of God.

The NET Bible places Mark 16:9-20 in double square brackets [[…]] to indicate that it is most likely an addition (a notation described in The Encyclopedia of New Testament Textual Criticism), and appends this note regarding the ending of Mark in different mss:

The Gospel of Mark ends at this point in some witnesses (א B 304 sys sams armmss Eus Eusmss Hiermss), including two of the most respected mss (א B). The following shorter ending is found in some mss: “They reported briefly to those around Peter all that they had been commanded. After these things Jesus himself sent out through them, from the east to the west, the holy and imperishable preaching of eternal salvation. Amen.” This shorter ending is usually included with the longer ending (L Ψ 083 099 0112 579 al); k, however, ends at this point. Most mss include the longer ending (vv. 9-20) immediately after v. 8 (A C D W [which has a different shorter ending between vv. 14 and 15] Θ Ë13 33 2427 Ï lat syc,p,h bo); however, Jerome and Eusebius knew of almost no Greek mss that had this ending. Several mss have marginal comments noting that earlier Greek mss lacked the verses, while others mark the text with asterisks or obeli (symbols that scribes used to indicate that the portion of text being copied was spurious).

The note then continues, offering the opinion that this evidence: “strongly suggests the secondary nature of both the short and the long endings.” However, as James Snapp Jr. points out in his blog:

The NET’s note contains some statements which are obvious mistakes, some statements which are probably erroneous, and some imprecise statements which give false impressions. The note also puts blinders over its readers' eyes by failing to mention some important evidence.

While the majority of extant mss of Mark end with the long ending (Mk 16:9-20) immediately following Mk 16:8, a number do not. Some end at Mk 16:8, several mss have both the short and long endings, and a few have some other slight variation on these endings, for example the inclusion of the Freer Logion between Mk 16:14 and 15 (referred to above by the NET as the "different shorter ending"). The majority view is that Mark originally ended at Mk 16:8, and that both the short and long endings are secondary, with both trying to ‘correct’ the problems introduced by the women failing to do in Mk 16:8 what they had just been told to do in Mk 16:7. In ‘A Textual Commentary on the Greek Gospels Vol. 2b The various endings of Mk,’ Wieland Willker summarizes the situation as follows:

The earliest evidence for the ending with ephobounto gar [i.e. ending at Mk 16:8] are the Gospels of Mt and Lk. Both follow Mk up to that point. After it they depart in very different ways. Attempts have been made to extract an ending, but these did not gain acceptance.
Comments from church fathers and introductory comments in the manuscripts indicate that many manuscripts ended with ephobounto gar in earlier times.
The origin of the shorter ending is only understandable, if the composer did not know the longer ending.
That the shorter ending always comes before the long one, seems to indicate a high respect (or a strong authority) for it. It is probable therefore that it is older than the long one.
The evidence indicates the existence of the long ending in the second half of the 2nd CE. So, probably also the short ending is as old as this.

The Shorter Ending of Mark

The shorter ending, found in L, Ψ, 083, 099, 274mg, 579, L1602, k, Sy-Hmg, samss, bomss, aethmss following Mk 16:8, reads:

And they reported all the instructions briefly to Peter's companions. Afterwards Jesus himself, through them, sent forth from east to west the sacred and imperishable proclamation of eternal salvation. Amen.

Willker (above) writes: “The origin of the shorter ending is only understandable, if the composer did not know the longer ending.” Where both the shorter and longer endings exist, the shorter always precedes the longer (suggesting the longer ending was added during copying to a ms that already contained the shorter ending), and there is only one extant ms that contains the shorter ending without the longer ending: Old Latin Codex Bobiensis (itk), about which Willker writes:

According to E.A. Lowe, k shows paleographical marks of having been copied from a second-century papyrus. Thus, the text of k is probably considerably older than k… The scribe of k appears to be very ignorant of Latin, probably a Copt, so it is certain that the ending is at least one copy older than k.

Itk is also notable because it contains changes to Mk 16:1-8 that suggest the author of the shorter ending was trying to ‘correct’ the problems caused by Mk 16:8 ending with ephobounto gar, in particular by removing Mk 16:8c: “neither said they any thing to any man; for they were afraid.” Itk also adds a reference to Jesus ascending between Mk 16:3 and 16:4, and Willker suggests that: “This drastic change points to a very early age, because only in the earliest times such major variations were possible.”

What Willker does not note is that adding the shorter ending after Mk 16:1-8 only makes sense if the author of the shorter ending only knew a version of Mark that ended at Mk 16:8. Based on the mss evidence, the only other version of Mark that the author could have seen would have included the longer ending, and there is no plausible explanation for anyone who knew of or made a copy of a ms of Mark that ended at Mk 16:20 replacing Mk 16:9-20 by the shorter ending. Even if the author of the shorter ending knew that Mark originally ended at Mk 16:8, and that Mk 16:9-20 was a later addition, there would be no reason to replace the longer ending by the shorter ending, or to insert the shorter ending between Mk 16:8 and Mk 16:9 (as seen in some mss). The addition of the shorter ending after Mk 16:8 only makes sense if the only ending of Mark known to the author was at Mk 16:8.

The question is then whether the author of the shorter ending had a ms of Mark that was clearly intended to end at what we see as Mk 16:8 (e.g. there was at least some space after Mk 16:8 in which more text could have been written), or one that only ended at that point because it was damaged. If the latter then it would be not unreasonable for a copyist to attempt to ‘fix’ the problem caused by the damage by adding a few words, such as found in the shorter ending. However, this on its own would not ease the problem caused by the text of Mk 16:8c conflicting with Mk 16:7, but with the knowledge that his exemplar was damaged anyway it is possible that the author of the shorter ending felt free to make other changes in the area of the damage. Later, when this ms was copied, there would be no evidence of the original damage at the end of Mk 16:8, and so some copyists may have ‘corrected’ the text by re-instating Mk 16:8c, while an un-corrected copy led to what we see in itk.

The alternative is that the copy of Mark used by the author of the shorter ending was not damaged, but instead ended neatly at Mk 16:8. In this case it seems likely that the author would attempt to ‘pick up the thread’ of Mk 16:8 at the beginning of his addition, but nevertheless in all but one ms that contains the shorter ending (itk) we still see a jarring transition, as noted by Willker:

It must be noted that the beginning of the shorter ending is in contradiction to the ending of verse 8:

16:8 "So they went out and fled from the tomb, for terror and amazement had seized them; and they said nothing to anyone, for they were afraid."

"But they reported briefly to Peter and those with him all that they had been told."

In verse 8 we are told that "they said nothing to anyone", but in the shorter ending they reported to Peter and those with him.

Although (above) Willker suggests that the text changes seen in itk indicate that this ms is very early, it is perhaps more likely that the changes to Mark 16 in itk are simply a result of the author knowing that his exemplar was damaged. In either event, the evidence suggests that the shorter ending is early. In addition, it is notable that this ending shows no knowledge of any additional appearances of Jesus to the disciples, suggesting that as well as preceding the longer ending, the shorter ending also precedes Matthew, Luke, and John.

The Long Ending of Mark

Why does the question of the ending of Mark matter? How Mark originally ended impacts what is known as the Synoptic Problem, which is: “an investigation into the existence and nature of the literary interrelationship among the three ‘synoptic’ gospels” (Mark, Matthew, and Luke). Depending on the synoptic theory, either Matthew and/or Luke were a source for Mark, or Mark was a source for Matthew and/or Luke. Either way, the question of how Mark ended is related to how both Matthew and Luke ended.

The final chapters of the synoptic gospels exhibit an interesting pattern. In each the early verses of their final chapter (Mk 16:1-8, Mt 28:1-8, and Lk 24:1-9) are broadly similar, and clearly depict the same events. Although some of the details vary, these three parallel accounts all describe a number of female followers of Jesus coming to his burial place, seeing the stone rolled away from the entrance, finding that Jesus’ body had gone, being told by one or two men (or an angel) that Jesus was risen, that (in Mark and Matthew) they would see him in Galilee, and then leaving. The three narratives then diverge beginning at Mt 28:9, Lk 24:10, and in Mark the gospel either ends here, or Mk 16:8 is followed by the short and/or long endings.

Each of these endings has very little in common with the others, and in ‘The New Testament in the Second Century: Text, Collections and Canon,’ Larry Hurtado suggests that the long ending of Mark is not original, but is instead related to the other gospels:

The “long ending of Mark,” as Kelhoffer has powerfully argued, is another striking textual phenomenon reflecting the four-fold Gospel tradition. This early addition to Mark appears to draw upon the four canonical Gospels, and no other gospel writing. It shows, too, that the four canonical Gospels were not only collected but also compared with one another, which explains best why someone thought that Mark’s ending was deficient and needed to be augmented along the lines of the other Gospels.

Also, in The authentication of Missionaries and Their Message in the Longer Ending of Mark Kellhoffer notes the references to Mary Magdalene in the long ending, but suggests that this may more likely reflect a connection to John rather than Matthew or Luke:

A more sure sign of literary dependence exists in the parallels to Jesus’ appearance to Mary Magdalene. The Gospel of John differs from Matthew and Luke in that it reports an appearance to Mary alone (John 20:11-18; cf. John 20:1) instead of to Mary and two other women (Mark 16:1-8; Matt 28:1-8) or to “certain women” (Luke 23:55-24:8). The name Μαριὰμ ἡ Μαγδαληνὴ could also have come from Matthew and Luke but probably reflects an allusion to John 20:18. The LE’s author also seems to have modified the participle ἀγγέλλουσα (John 20:18) to ἀπήγγειλεν in Mark 16:10, perhaps in light of the indicative form (ἀπήγγειλαν) in Matthew 28:8-11 and Luke 24:9. As observed in the statistical analysis of chapter 2, a particularly distinctive feature John and the LE share is how they use the pronoun ἐκείνη to denote “that woman.” Such a use of ἐκείνη most probably stems from a deep familiarity with this Johannine passage.

Kellhoffer also makes note of the fact that Mary Magdalene is introduced in Mk 16:9 as if there had been no previous mention of her in Mark:

Now when Jesus was risen early the first day of the week, he appeared first to Mary Magdalene, out of whom he had cast seven devils. [Mk 16:9]

However, Mary Magdalene is referred to by name in Mk 15:40, 47 and 16:1, so there should have been no reason to introduce her again in Mk 16:9. In addition, the detail about the seven devils (otherwise only seen in Lk 8:2) would most naturally have been included in Mk 15:40, where Mary is first named. This, together with the fact that Mk 16:9 ignores what had just been stated in Mk 16:8, strongly suggests that Mk 16:9 was written as the introduction to a piece of text that was never originally intended to follow Mk 16:1-8.

Willker (above), notes that the authors of Matthew and Luke show no signs of having seen Mk 16:9-20, while in ‘The Authenticity of Mark 16:9-20’ (Summary available here) Snapp presents evidence for the corollary, namely that the author of Mk 16:9-20 had not read either Matthew or Luke:

● The Author of 16:9 to 20 Had Not Read Matthew.

Mark 16:9 through 11 states that Jesus appeared to Mary Magdalene and she reported this to the disciples, but they did not believe her. Matthew 28 gives no indication that the women’s report about the appearance of Jesus was not believed by the disciples: in Matthew 28, verse 11, the women continue on their way to the disciples, and in Matthew 28, verse 16, the disciples have accordingly gathered in Galilee. The author of Mark 16:9 to 20 has not simply repeated material in Matthew; on the contrary, he has recorded an event in verse 11 (the eleven apostles’ disbelief in Mary’s report that Jesus was alive and that she had seen Him) which no one would naturally derive from the Gospel of Matthew.

● The Author of 16:9 to 20 Had Not Read Luke.

In Luke 24:1 through 11 and Luke 24:22 through 24, the women encounter angels, not Jesus. The author of Mark 16:9 to 11, if he relied on the Gospel of Matthew, had no basis to write that the women’s report had not been believed. Yet, if he had depended on the Gospel of Luke, he had no basis to report that Mary Magdalene had seen Jesus.

Also, Mark 16:14 states that “Later He appeared to the eleven as they sat at the table, and He rebuked their unbelief and hardness of heart, because they had not believed those who had seen Him after He had risen.” There is no such rebuke in Matthew. In Luke, when Jesus appears to the eleven disciples, He gives no such rebuke. The mild questions in Luke 24:38 focus on the disciples’ hesitance to believe their eyes and ears, not their disbelief of the report of earlier eyewitnesses to His resurrection.

In addition, Luke 24:36 presents Jesus’ appearance to the eleven right after the two Emmaus-road travelers arrive and tell about their experience. Luke does not say that the eleven disciples disbelieved their report. Luke does not even make it clear that the eleven disciples had time to do so before Jesus personally appeared to them. No writer, having read Luke 24:33 through 43, would summarize it as two events in which the disciples first rejected the report of the two travelers, and were later rebuked by Jesus for doing so. A harmonization of the two accounts is achievable by positing that the disciples’ conversation lasted a long time, but it is unlikely that anyone would create a text that required such a harmonization when other options were available.

Also, Luke locates Christ’s ascension at the Mount of Olives (Luke 24:51, Acts 1:9). The author of Mark 16:9 to 20, if he had written in order to compose an ending for the Gospel of Mark, and if he had done so with the Gospels of Matthew and Luke on hand, would have possessed a very strong impetus to mention that the disciples traveled to Galilee and saw Jesus there, and then returned to Jerusalem where He ascended to heaven. Yet there is no statement anywhere in Mark 16:9 to 20 that the disciples, last seen in Jerusalem, traveled back to Galilee.

Snapp then presents similar evidence indicating that the author of Mk 16:9-20 had not read John or Acts either, and then states:

As Hort perceived, 16:9 to 20 is “apparently older than the time when the Canonical Gospels were generally received; for, though it has points of contact with them all, it contains no attempt to harmonise their various representations of the course of events.”

If, as the evidence strongly suggests, Mk 16:9-20 was written without any knowledge of Matthew, Luke, John, or Acts, then it seems almost certain that it was written before them. On the other hand, if so, then why do Matthew and Luke show no sign that their respective authors (aMatthew and aLuke) knew Mk 16:9-20? As indicated above, Hort noted that Matthew and Luke have “points of contact” with Mk 16:9-20, but this is far short of any suggestion that either aMatthew or aLuke knew what we see as Mk 16:9-20. Snapp suggests that this is the effect “of reporting some of the same events,” but it could also have resulted from aMatthew and aLuke having both seen a document containing text parallel to that at the points of contact, and related to but shorter than Mk 16:9-20.

Depending on the assumed synoptic hypothesis the similarities and differences between the final chapters of Matthew and Luke are each either easy or difficult to explain. For example, theories in which neither aMatthew nor aLuke saw the other’s gospel, such as the Two Source Hypothesis (2SH) or Mark-Q theory, have no problem with the differences but require another source used by both authors to explain the similarities. Conversely, theories in which either aMatthew or aLuke did see the other’s gospel, such as the Farrer Theory or Mark Without Q Hypothesis (MwQH), do not need another source to explain the similarities but instead have difficulty explaining why the resurrection accounts in Matthew and Luke are so different.

On the Mark-Q theory both aMatthew and aLuke knew and used Mark and an otherwise unknown source, Q, that among other things is defined as not containing a resurrection account. If Mark as seen by aMatthew and aLuke ended at Mk 16:8 (i.e. did not contain a resurrection account) then we would not expect to see any similarities between any of the later endings of Mark, and the endings of either Matthew or Luke. In contrast, if Mark already contained the long ending, then both aMatthew and aLuke would have seen it and we would then expect to at least see parallels to the long ending in both Matthew and Luke.

Although the resurrection accounts in Matthew and Luke are very different, there are a few places where they do appear to parallel the long ending, in particular what is known as the Great Commission at Mk 16:15-16, which has parallels at Mt 28:18-20a and Lk 24:46-47. Even so, the great majority of the long ending of Mark has very little in common with Mt 28:9-20, and although Lk 24:10-53 does contain parallels to more of the long ending than does Matthew it also contains a lot of material unique to Luke. Although this could suggest that aMatthew and aLuke each took a small part of the material from the long ending of Mark and expanded upon it, instead as indicated above it is very unlikely that either saw Mk 16:9-20. This is a problem for the Mark-Q theory, in which by definition both aMatthew and aLuke would have seen and used Mark but could not have known the long ending, and so must have obtained the text that is common to their resurrection accounts from some other source.

This particular point is not a problem on the MwQH, as in this case aLuke knew both Mark and Matthew, and so he could have seen the Great Commission in Matthew and created a version of it himself. It is also not a problem in any general synoptic hypothesis that has a second source (SS) for Matthew and Luke (i.e. in addition to Mark) in which SS is allowed to contain a resurrection account, and aLuke is allowed to know Matthew (The Mark with Second Source, or MwSS, hypothesis). If neither Mark nor SS contained a resurrection account, then, as on the MwQH, the parallels between the resurrection accounts in Matthew and Luke are still explainable as a result of aLuke knowing Matthew. However, on the MwQH the fact that aLuke knew Matthew makes it hard to explain why he apparently ignored most of the ending of Matthew, from exactly the point where the parallels to Mark end. Things are different on the MwSS hypothesis, because SS, which aLuke also knew, could have contained a resurrection account different to all of those that we see in Mark, Matthew, and Luke, but containing the “points of contact” referred to above.

One of the issues that is not normally considered in discussions of the resurrection accounts is the group of shorter textual variants (mainly found in Lk 24) that are known as Western Non-interpolations (the name Non-Western Interpolations is not used although it is potentially just as valid). As the MwQH, in common with most synoptic theories, does not consider textual variants it does not address the differences between canonical Lk 24 and the significantly shorter Western text, in which the following are not present:

“of the Lord Jesus;” [Lk 24:3]

“He is not here, but has been raised!;” [Lk 24:6]

“from the sepulchre;” [Lk 24:9]

The whole verse regarding Peter; [Lk 24:12]

“and said to them, ‘Peace be with you;’” [Lk 24:36b]

“When he had said this, he showed them his hands and his feet;” [Lk 24:40]

“and was taken up into heaven;” [Lk 24:51]

“worshiped him and.” [Lk 24:52]

Because in the MwSS hypothesis SS is one of the sources of canonical Luke it is conceivable that SS had a resurrection account that was shorter than that in canonical Luke (specifically not including the words shown above), that was then seen by aLuke, and mss evidence of which has only survived in D and the Old Latin. However, if SS contained everything that we see in the Western text of Lk 24:10-53, then on the MwSS hypothesis aMatthew would have also seen this text, and we would therefore expect most, if not all, of these verses to have parallels in Matthew. Because they do not the most likely scenario would be that as well as not containing the text of the Western Non-Interpolations, SS only contained just a small part of what we see in Lk 24:10-53, most likely excluding much of the text that has no parallel in Matthew.

It is notable that in Luke the narrative describing what the disciples did after finding the empty tomb is twice ‘interrupted’ (in the sense that the narrative works perfectly well without it) by two meetings with Jesus, most of which is unique to Luke. In the first meeting (Lk 24:16-32) Jesus pretends to be someone else and recaps for the disciples everything foretold about him by the OT prophets, while in the second (Lk 24:35b, 36b-45, 46b) he makes it as plain as possible (to both the disciples and us, the readers of Luke) that he had a physical body, before referring back to the OT again. Finally, in Lk 24:47b-49 Jesus hints at what we see in Acts 2 (suggesting perhaps that this is a later addition to Luke) when he tells the disciples to wait in Jerusalem until they receive power. If none of these ‘interruptive’ passages were in SS then the remaining text (using Lukan verse numbering) would read approximately as follows (using the KJV translation):

It was Mary Magdalene and Joanna, and Mary the mother of James, and other women that were with them, which told these things unto the apostles. [24:10]

And their words seemed to them as idle tales, and they believed them not. [24:11]

And, behold, two of them went that same day to a village called Emmaus, which was from Jerusalem about threescore furlongs. [24:13]

And they talked together of all these things which had happened. [24:14]

And it came to pass, that, while they communed together and reasoned, Jesus himself drew near, and went with them. [24:15]

And they returned to Jerusalem, and found the eleven gathered together, and them that were with them, [24:33]

And they told what things were done in the way. [24:35a]

And as they thus spake, Jesus himself stood in the midst of them [24:36a]

And said unto them, [24:46a] that repentance and remission of sins should be preached in his name among all nations, [24:47a]

And it came to pass, while he blessed them, he was parted from them. [24:51]

And they returned to Jerusalem with great joy: [24:52]

And were continually in the temple, [praising and][blessing] God. [24:53]

This shortened text bears a great similarity to the long ending of Mark, with all but one verse of the long ending having a parallel in this shortened account, and three verses also having parallels in Matthew:

Mk 16:9-11: A repetition of the fact that Jesus appears to Mary Magdalene. She then tells an unspecified number of followers who do not believe her (c.f. Lk 24:10-11).
Mk 16:12-13: Jesus appears (as someone else) to two followers as they are walking somewhere. They tell the remaining followers, who do not believe them (c.f. Lk 24:13-15, 33, 35a, Mt 28:9a).
Mk 16:14: Jesus appears to the eleven as they are eating, and complains about them not believing those who had seen him (c.f. Lk 24:36a).
Mk 16:15-16: The Great Commission (c.f. Lk 24:46a, 47a, Mt 28:18-19).
Mk 16:17-18: Jesus describes the signs that believers will receive (c.f. Lk 24:49 – This may be a later addition to Lk).
Mk 16:19-20: Jesus is taken away (c.f. Lk 24:51-53)

On this basis it appears that not only is what we see in Lk 24:10-53 an expansion of what was in SS, but also the long ending of Mark was derived from the ending of SS, and was added to Mark after SS was written.

Conclusion

As indicated above, the evidence strongly suggests that neither aMatthew nor aLuke knew a version of Mark containing the long ending (Mk 16:9-20). However, the evidence is just as strong that the author of the long ending did not know either Matthew or Luke. Nevertheless, some of the elements in the long ending do have parallels in Luke (and to a much lesser extent Matthew), suggesting that both aMatthew and aLuke knew something that had at least some commonality with the long ending. Although most synoptic theories do not allow for such a ‘resurrection’ source prior to either Matthew or Luke (other than Mark itself), on the MwSS hypothesis SS fulfills this role. Although not all the details in the long ending of Mark are identical to those in the ending of SS suggested above, the parallels between the two are at least as close as other well established synoptic parallels, strongly suggesting a literary relationship between them.

As also reported above, the long ending of Mark is unlikely to have been original, so suggesting that it derives from the short resurrection account in SS rather than the other way round. On the MwSS hypothesis aMatthew would also have seen this short account in SS, but with no resurrection account in his primary source, Mark (which ended at Mk 16:8), he chose to largely reject that in SS. Instead, he created his own unique account at Mt 28:9b-16, although he did create three parallels to the text in SS: Jesus appearing first to the two Marys; then later to the eleven; and finally the Great Commission. It should be noted that the same applies to the MwEL hypothesis, because in this hypothesis an early version of Luke (Early Luke) is a special case of SS (see Luke Chapters 1 and 2 for evidence in Luke itself that an early version of Luke did exist).