Where Did Jesus Go? – Critical Notes on the Ascension, Pt 1

This post is an extended “Note of the Day” following the celebration of the Ascension of Jesus, traditionally commemorated 40 days after Easter (cf. Acts 1:3ff). There are surprisingly few direct references to the “Ascension”, either in the New Testament or in early Christian literature. More commonly, reference is made to Christ’s exaltation (usually involving either the verb u(yo/w, “to raise high” or adjective u(yhlo/$, “high”; cf. Acts 2:33; 5:31; Philippians 2:9; Hebrews 1:3; 7:26, etc), or to his being in heaven at the “right (hand)” (decio/$, that is, the giving/receiving hand) of God (cf. Mark 14:62 par.; Acts 2:33-34; 5:31; 7:55-56; Romans 8:34; Ephesians 1:20; Colossians 3:1; Hebrews 1:3, 13; 8:1; 10:12; 12:2; 1 Peter 3:22).

Only in Luke-Acts is an ascension described as a distinct event, perceivable in space and time. However, since there are numerous difficult text-critical (and interpretative) issues related to these passages, these will need to be discussed in some detail.

Luke 24:50-53

Luke’s Gospel concludes with a scene (apparently still on Easter day) which, in the “oldest and best” manuscripts (Ë75 a B C* L 1 33 579 etc), reads as follows:

50  )Ech/gagen de\ au)tou\$ [e&cw] e%w$ pro\$ Bhqani/an kai\ e)pa/ra$ ta\$ xei=ra$ au)tou= eu)lo/ghsen au)tou/$. 51 kai\ e)ge/neto e)n tw=| eu)logei=n au)to\n au)tou\$ die/sth a)p’ au)tw=n kai\ a)nefe/reto ei)$ to\n ou)rano/n.
“And he brought/led them out[side] until toward Bethany, and lifting over (them) his hands he spoke well to them [i.e. blessed them]; and it came to be, in his speaking well to them [i.e. blessing them], he stood (apart) from them and was carried up into the heaven.”
52 Kai\ au)toi\ proskunh/sante$ au)to\n u(pe/streyan ei)$  )Ierousalh\m meta/ xara=$ mega/lh$ 53 kai\ h@san dia/ panto\$ e)n tw=| i(erw=| eu)logou=nte$ to\n qeo/n.
“And they, kissing toward him [i.e. worshiping him], turned back unto Jerusalem with great joy, and they were through all [i.e. continually] in the sacred place [i.e. temple] speaking well to [i.e. blessing] God.”

(The Majority text differs slightly, primarily in reading ei)$ Bhqani/an instead of pro\$ Bhqani/an in v. 50, and adding kai\ ai)nou=nte$ or ai)nou=nte$ kai\ [“blessing and praising God”] in v. 53.)

There are, however, two major variants (omissions) in the key Western MSS (D, Old Latin a b d e ff2 l, and the Sinaitic Syriac):

    1. Verse 51 reads: kai\ e)ge/neto e)n tw=| eu)logei=n au)to\n au)tou\$ die/sth a)p’ au)tw=n “and it came to be, in his blessing them, he stood (apart) from them” (without kai\ a)nefe/reto ei)$ to\n ou)rano/n “and he was carried up into the heaven”). In other words, it relates that Jesus simply “parted” from them, without any reference to an ascension into heaven.
    2. Verse 52 continues: kai\ au)toi\ u(pe/streyan ei)$  )Ierousalh\m meta/ xara=$ mega/lh$ “and they turned back unto Jerusalem with great joy…” (without proskunh/sante$ au)ton “worshiping him”).
      See how this shorter version of vv. 50-53 reads, in context, in conventional translation:
      “And he led them out toward Bethany, and raising his hands over (them) he blessed them; and it came to be, in his blessing them, (that) he parted from them; and they returned to Jerusalem with great joy, and they were continually in the Temple, blessing God.”

These are both so-called Western “Non-Interpolations”, i.e. instances where the shorter reading of the (usually longer and more expansive) Western text has been thought (by some scholars) to preserve the original reading in the face of superior manuscript evidence (I have discussed the other seven key “Non-Interpolations” in a previous post). The first of these two (in v. 51) is far more significant, especially since, in addition to the Western MSS, the shorter reading is also found in the Georgian version (group 1) and the original hand of Codex Sinaiticus (a*).

How is one to explain this variant? As indicated above, the vast majority of MSS, including all the early/best Greek MSS (Ë75, a [corrected], A, B, C, K, L, W, X, D, etc.) contain the words kai a)nefe/reto ei)$ ton ou)rano/n. The manuscript evidence would seem to be decidedly in favor of the longer reading, but internal considerations make it a bit less certain. In which direction did the change occur? There are a number of possibilities:

Reasons for Omission (in support of the longer text):

  1. To avoid contradiction with the chronology in Acts. It is certainly possible that scribes, noticing the apparent discrepancy between v. 51 and Luke’s own account of the Ascension in Acts 1:1-11, deleted the words. In the Gospel, it would seem that the Ascension takes place on the same night as the Resurrection, whereas in Acts (v. 3) it occurs 40 days later. This is probably the most popular explanation.
  2. A scribal mistake. A scribe may have skipped from a)p’ au)twn kai in v. 51 to ou)ranon kai au)toi at the end of v.51 & start of v. 52 (homoioarcton: each has the segment nkai). However, this would require that (the precursors of) a and D both made the same mistake, which is rather unlikely.
  3. Theological reasons. Some scholars have thought that the so-called “Non-Interpolations” (involving the Resurrection appearances and “Ascension”) exhibit a purposeful tendency in the Western text (in Luke-Acts) to eliminate concrete references to the resurrection body of Jesus, and physical nature of the Ascension, etc. With regard to the Ascension in particular, see especially Eldon J. Epp’s article “The Ascension in the Textual Tradition of Luke-Acts”, in New Testament Textual Criticism: Its Significance for Exegesis. Essays in Honor of Bruce M. Metzger. Oxford: Clarendon, 1981, pp. 131-145.
  4. The support of Acts. Acts 1:2 would seem to indicate that the Gospel referenced the Ascension (a&xri h!$ h(me/ra$a)nelh/mfqh, “until which day…he was taken up”). Assuming this is the case, it could be (rightly) argued that the author would not say he described an event which he in fact did not record. It should be noted that several Western witnesses (gig, quotations in Augustine and Vigilius) also omit reference to the ascension (a)nelh/mfqh) in this verse.

Reasons for Addition (in support of the shorter text):

  1. Literary or Theological reasons. Although Luke-Acts may have been published together as a ‘two-volume’ work, by the mid-second century (at the latest), the Gospel of Luke was being copied and distributed bound together (in codex form) with the other Gospels; meaning that, as in nearly all printed New Testament editions today, it was separated from the book of Acts. The shorter reading, if original, would close the Gospel with the suggestion that Jesus simply “parted” from the disciples—a rather unexciting and possibly misleading conclusion. The scribal tendency was always to add Christological details, rather than remove them; it would have been natural to add the few extra words (both in v. 51 and 52), in order to exalt the portrait of Christ.
  2. The shorter text removes the chronological difficulty with Acts. This argument cuts both ways (see above), for the longer text could be said to be the more difficult reading (lectio difficilior potior). However, since Luke explicitly records the Ascension taking place at least 40 days after the Resurrection (Acts 1:3ff), would he (the same author of Luke-Acts, by general consensus) have created the confusion by recording the Ascension (apparently) taking place on the day of the Resurrection (Luke 24:50-53)?
  3. Additional support from Acts. It is possible that the phrase a&xri h!$ h(me/ra$a)nelh/mfqh (“until which day…he was taken up”) in Acts 1:2 should not be taken to imply that the Ascension was narrated in the Gospel, but only events which took place prior to that day. In this regard, to note the reference (v. 22) in Peter’s subsequent address (Acts 1:15-22), where nearly similar language is used. Could the author of Acts simply be reproducing the phrasing from v. 22, as part of his “prologue”, without specific reference to details in the Gospel?
  4. Evidence from the Church Fathers. The Ascension is referred to numerous times in writings of the 1st-3rd centuries, for example: Epistle of Barnabas 15; JUSTIN: 1 Apology 26, Dialogue with Trypho 82, 87, On the Resurrection ch. 9; IRENAEUS: Against Heresies I.10, III.17, IV.33.13, 34.3, V.31, Demonstration of the Apostolic Preaching 41, 84; CLEMENT: Stromateis VI. ch. 15; ORIGEN: On First Principles Pref §1, II.6.1, 7.2, On Prayer §23, Against Celsus VII.8; TERTULLIAN: Against Marcion V.8, Against Praxeas 25, 30, Prescription Against Heretics 13, On the Resurrection 51; The Muratorian Canon; Epistle of the Apostles 18; Cyprian On the Lord’s Prayer §8, etc. (by no means an exhaustive list). Most of these references are to the narrative in Acts 1:9ff; Ephesians 4:9-10, or to the belief generally; however, I have not been able to find a single clear reference to the long text of Luke 24:51-52 cited in any writing up through the third century (outside of the Diatessaron [§55], a work with a singularly difficult textual history). Moreover, in Tertullian’s fourth book Against Marcion, in which he goes over many details of Luke’s Gospel, up through the Resurrection appearances (chapter 43), he does not cite the long text of v. 51 or 52, and makes no reference to the Ascension (cf. Bart Ehrman, The Orthodox Corruption of Scripture, pp. 257-8).
  5. The Western Non-Interpolations. Despite protests from scholars on both sides of the argument, it is hard to avoid the notion that the 9 key “non-interpolations”, eight of which are all found together in the same set of MSS (D a b d e ff2 l), stand or fall together—most likely, they are all original, or they are not. If one accepts the shorter text in the previous 7 Lukan instances, then one really ought to do so here as well.

Clearly, intrinsic/transcriptional arguments can be made for sides. Ultimately, it is difficult to ignore the overwhelming textual evidence. If the longer reading is, in fact, original, I suspect that the apparent discrepancy (with Acts) may be the result of Luke compressing/conflating the narrative, thereby giving the impression that it all happened on one night. This sort of handling of historical narrative was quite common with ancient writers, as unsatisfying as it might be to our modern sensibilities. On the other hand, the clear scribal tendency was to add significant Christological details to the Gospel narrative, rather than omit them (even when there are apparent discrepancies involved); it seems to have been much more acceptable to modify (instead of deleting) difficult words in the text. The presence of the longer reading(s) in the Bodmer Papyrus (Ë75, c. 200) have turned the tide decisively; however, I am by no means so certain the shorter reading(s) can be dismissed as easily as many commentators do today.

The Resurrection in Luke: The ‘Western’ Text

The major text-critical question in the Resurrection Narratives involves the so-called “Western Non-Interpolations” in the Gospel of Luke. This rather awkward term stems from the analysis by Westcott & Hort (principally Hort) in their landmark The New Testament in the Original Greek (1881, vol. II pp.175-177), regarding situations where, despite superior manuscript evidence to the contrary, the Western Text may have the original reading. In general, the “Western Text” (as represented by Codex Bezae [D], key Old Latin [and Old Syriac] MSS, and other versional witnesses), was deemed inferior to the so-called “Neutral Text” (exemplified esp. by Codex Vaticanus [B])—this view, with some modification (and different language), continues to be held by most critical scholars today. Particularly in Luke-Acts, the “Western Text” tends to have longer readings at key variation-units—expanding or adding clarifying detail to the text. It is all the more noticeable, then, on those rare occasions when D (and other Western witnesses) happen to contain a shorter reading. When this fact (cf. the principle lectio brevio potior, “the shorter reading is preferrable”) is combined with intrinsic or transcriptional probability in favor of the shorter text, one must then contend with the possibility that the Western reading is original. Hence the term “Western Non-Interpolation”: i.e., the majority text contains an interpolation (an added verse or phrase), contrary to the shorter (original) Western text.

Westcott & Hort identified 27 shorter Western readings of note: six were deemed unlikely to be original, twelve others considered possibly (but probably not) original, and nine regarded as “probably original”. These nine (the “Non-Interpolations”) are: Matthew 27:49; Luke 22:19b-20; 24:3, 6, 12, 36, 40, 51, 52. For some time, critical scholars tended to favor this approach; however, in recent decades, with the discovery of the Bodmer Papyri (esp. Ë75), the pendulum has swung decidedly in the opposite direction—the majority of scholars, on the whole, now reject these shorter Western readings. Indeed, Ë75 (early 3rd century?) contains the longer (majority) reading for all 8 Lukan “Non Interpolations”, greatly strengthening the already impressive external evidence for them. On the other hand, the strongest argument in favor of the shorter readings is one of transcriptional probability—no one has really been able to offer a good explanation as to how (or why) the longer readings, if original, would have been deleted. Moreover, nearly all of the majority readings in these instances involve (possible) harmonizations to other portions of the New Testament (see notes below) as well as significant Christological details, both of which are more likely to represent scribal additions than details scribes would have ever deleted. For a fairly thorough defense in favor of the Lukan “Non-Interpolations”, see B. Ehrman, The Orthodox Corruption of Scripture Oxford:1993, pp. 197-232.

There is the problem: on the one side, the external manuscript evidence is decidedly in favor of the longer readings; but on the other, internal transcriptional evidence seems clearly to favor the shorter. Interestingly, all of the nine “Non-Interpolations” are from the Passion and Resurrection narratives (8 from the Lukan), and all but two (7) from the Resurrection/Ascension accounts in Luke 24 (common to virtually the same set of manuscripts). This cannot be coincidental, nor do I think it can be accidental. In other words, whichever set of readings (longer/shorter) is correct, the changes seem to have been both deliberate and consistent in Luke 24. Either scribes added text (interpolations), perhaps to harmonize with John’s account (see below) etc. and/or enhance the Christological portrait, or they deleted the text, for reasons that are as yet not entirely clear.

Luke 24:3

Here is a translation of the majority text of vv. 1-4, with the words in question italicized:

1And on (day) one of the week, of deep dawn [i.e. early at dawn], upon the memorial [i.e. tomb] they came carrying spices which they had made ready. 2And they found the stone having been rolled (away) from the memorial, 3but going into (it) they did not find the body of the Lord Jesus. 4And it came to be in their being without a way-through [i.e. as they were at a loss] about this, and see!—two men stood upon [i.e. next to] them in flashing clothes…

Manuscripts D a b d e ff2 l r1 do not include the words tou= ku/riou  )Ihsou=. They may have been added to specify and make clear what would otherwise be implied: that it was truly Jesus’ body missing from the tomb. If the words did not drop out by accident, it is hard to explain why a scribe (on orthodox one, at least) would have removed them. A few manuscripts (579 1241 pc syrs, c, p bohms) read simply tou=  )Ihsou=.

Luke 24:6

The same group of Western manuscripts (along with Georgian MS B) do not include the words ou)k e&stin w!de a)lla\ e)ge/rqh from the angelic announcement. Here is a translation of the majority text (with italicized words):

5And at their [i.e. the women] coming to be afraid and bending th(eir) faces into the earth, they [i.e. the men/angels] said to them, “(For) what [i.e. why] do you seek the living amid the dead? 6He is not here, but he has risen! Remember how he spoke to you…”

Luke 24:12

Almost the same group of Western MSS (along with several Syriac witnesses [and Marcion?]) do not include verse 12 at all. The majority text reads:

o( de\ Pe/tro$ a)nasta\$ e&dramen e)pi\ to mnmei=on kai\ paraku/ya$ ble/pei ta\ o)qo/nia [kei/mena] mo/na, kai\ a)ph=lqen pro\$ e(auto\n qauma/zwn to\ gegono/$

But Peter, standing up, ran upon [i.e. ran to] the memorial [i.e. tomb] and bending alongside he saw the cloths [laying] alone, and he went from (there) toward his own (home), wondering at the (thing which) had come to be [i.e. what had happened]

This is of course quite similar to the account in John 20:4-5f, enough that scholars who favor the shorter reading view the verse as a harmonizing interpolation. The word kei/mena (not in Ë75 a B W etc) is probably a simple harmonization; however, otherwise, there are enough differences (including all of 12b), that this is less likely for the verse as a whole. On the other hand, the sequence from verse 11 to 13 reads smoother without v. 12:

11and these words [i.e. the women’s report] shined in their face [i.e. appeared to them] as if idle-talk, and they [i.e. the apostles] did not trust them [i.e. the women]. 13And see—two of them [i.e. disciples/apostles] in the self(-same) day were traveling unto a village…

It is also much more effective dramatically without v. 12, leading up to the revelation at Emmaus; it can be argued that the announcement in v. 34 (“the Lord has been seen by Simon!”) is more dramatic this way as well. That being said, what of the (internal) evidence—the intrinsic or transcriptional probability—for inclusion/exclusion of the verse? I find the argument for simple harmonization with John to be weak; I am also unconvinced by the idea that the verse was added to make better sense of v. 34. A much stronger argument is that the verse was added (whether from John, or more likely a separate tradition) to soften the image of the unbelieving apostles in v. 11—not all of them mistrusted the women, Peter responded aggressively to see for himself! What of reasons for scribes’ deleting the verse? Apart from the fact that the narrative reads better without v.12 (the plural pronoun and copulative kai arguably connect more readily with v.11), it is hard to come up with a good explanation.

Luke 24:36

Here the opening of Jesus’ introduction—kai\ le/gei au)toi=$: ei)rh/nh u(mi=n—is not included by the same group of Western manuscripts (D a b d e ff2 l r1). Again, let us examine the context in translation (disputed words italicized):

36And as they spoke this, (Jesus) himself stood in the middle of them and says to them: “Peace to you”. 37But being terrified and coming to be in fear, they seemed to gaze at a ‘spirit’. 38And he said to them, “(For) what [i.e. why] are you disturbed…?”

The scene makes more immediate sense without the words—Jesus suddenly appears in their midst and they are terrified (presumably not recognizing him, cf. v. 16ff). There would seem to be less reason for such sudden, extreme fear, after the words of greeting (“Peace to you”). In this instance, a harmonization with John (20:19) is perhaps more likely than in Luke 24:12. As for omission, if the words did not fall out accidentally, why would they have been deleted? Again, it is hard to come up with a reason.

Luke 24:40

Here, as at 24:12, and entire verse is missing from (the same group) of Western manuscripts, along with the Curetonian and Sinaitic Syriac. The verse reads:

kai\ tou=to ei)pw\n e&deicen au)toi=$ ta\$ xei=ra$ kai\ tou\$ po/da$
“and having said this, he showed them the hands and the feet”

A harmonization with John 20:20 is certainly possible. On the other hand, I would say that there is at least a plausible reason for scribes omitting the words, as they may have appeared superfluous or redundant directly following v. 39.

Luke 24:51-52

These two variations units are, in some ways, even more controversial, and are better left to an (upcoming) article on the Ascension.

One of the reasons earlier scholars more readily favored the “Non-Interpolations” of vv. 12, 36, and 40, was the understandable assumption that these were scribal harmonizations (of a sort all too common in the manuscripts) with the parallel passage in John. However, commentators today tend to prefer the view that Luke and John (in the Passion and Resurrection narratives, at least) both draw from a common tradition, which explains their sharing certain details not found in Matthew-Mark.

From a text-critical point of view, however, it should be reiterated that the internal evidence favors all of the Lukan “Non-interpolations” (in chapter 24). The two overriding arguments:

    1. Scribes are more likely to have harmonized the text (to another Gospel passage) by adding to it, than to eliminate a harmonization by deleting the text.
    2. Scribes are more likely to add details enhancing or expanding the portrait of Christ, than to delete them. One indisputable fact is that for all seven instances in Luke 24, the longer (majority) text adds vivid or significant detail related to the reality of Jesus’ resurrection not found in the corresponding Western text.

All things considered, it is safest to defer to the overwhelming external evidence in favor of the longer readings. Yet, in studying and meditating upon the Resurrection accounts in Luke, I would urge care and consideration—if we wish to understand the inspired original text, such significant textual variants must be given their due.