What Roth and BeDuhn Left Open
    Canon & Text
    Scholarship

    What Roth and BeDuhn Left Open

    November 5, 2026

    Two of the most cited modern reconstructions of Marcion’s Gospel, Dieter Roth’s and Jason BeDuhn’s, decline to rule on a large part of Luke. By our count, Roth leaves 385 of Luke’s verses undecided, and BeDuhn 301. A reader of Marcion’s Gospel still has to be told something about each of those verses. Here is what my Critical Edition of the Greek True Testament says about every one of them, what the church fathers and the German editors before Roth said, and where the evidence runs out for all of us.

    Two Careful Reconstructions

    Jason BeDuhn’s The First New Testament (2013) and Dieter Roth’s The Text of Marcion’s Gospel (2015) are careful books. Both use a third category beside present and absent: unattested, for verses on which they find the evidence too thin to decide.

    That discipline is right as far as it goes, and I share it: not attested is not the same as absent. But it leaves a great deal open. By our count of their two books:

    • Roth marks 385 verses of Luke unattested, a third of the Gospel.
    • BeDuhn marks 301.
    • 280 are left open by both, and 406 by at least one of them.

    Not every editor left them open. August Hahn’s reconstruction of 1823 prints Greek for 334 of Roth’s 385, and Matthias Klinghardt counts 308 of them in Marcion’s Gospel. Harnack and Zahn ruled on many of them one by one, sometimes in, sometimes out, and on others they too wrote that nothing is known.

    My Critical Edition of the Greek True Testament, the Greek text of Marcion’s Bible with the evidence for every word set out beside it, takes the harder road. It gives every one of these verses a label, and for most of them a stated reason, citing the fathers and the German editors where they speak. This article follows Roth’s 385 through the Critical Edition. BeDuhn’s 301 come out the same way, and his figures are given at the end.

    Where Roth’s 385 Verses End Up

    • 28 have a witness. I count them attested because Tertullian reports the scene.
    • 58 are kept as probable. No witness quotes them, but verses that are attested cannot stand without them.
    • 73 are judged absent. For 4 of them a Greek witness reports the absence. For 69 the absence is probable, for reasons the Critical Edition gives.
    • 226 remain open. I agree with Roth that no one can say, and the Critical Edition marks them so.

    So I decide 159 of Roth’s 385 verses, and I agree with him that the other 226 cannot be decided. The difference is that the Critical Edition says so for each one, in print.

    Twenty-Eight Verses With a Witness

    For 28 of Roth’s unattested verses, I count Tertullian’s Against Marcion as reporting the scene. He does not quote the Greek, so the Critical Edition prints Nestle’s wording in single brackets, but a reported verse is attested. BeDuhn counts 26 of these 28 in Marcion’s Gospel, and Klinghardt 27.

    • Luke 9:52-53, the Samaritan village. Tertullian reports the village of the Samaritans (Marc. 4.23.8) and its refusal (4.29.12): “Jesus sent messengers ahead of him, and when they went, they entered a Samaritan village to get things ready for him. But the people there would not welcome him, because he was heading for Jerusalem” (Evangelion 6:9-10 TT).
    • Luke 20:22-23, the question about taxes. Tertullian writes of “the consultation about the tribute: And the Pharisees came to him, testing him” (de tributi consultatione, Et accesserunt ad eum pharisaei temptantes eum, 4.19.7). His words are Matthew’s, so they show only that the question was put to Jesus in Marcion’s Gospel, not its wording: “They asked him, ‘Are we permitted to pay taxes to the emperor, or not?’ But Jesus perceived their craftiness, saying to them,” (Evangelion 13:10-11 TT).
    • Luke 4:33, the demon in the synagogue. Tertullian introduces the demon’s words with “there the spirit of a demon cries out” (Exclamat ibidem spiritus daemonis, 4.7.9): “In the synagogue there was a man possessed by a demon, and he suddenly shouted out in a loud voice,” (Evangelion 1:4 TT).
    • Luke 19:12-21, the parable of the pounds. Eight of the 28 are here. Tertullian calls it “the parable of the servants” (Servorum quoque parabola, 4.37.4). The case of Luke 19:12 is set out in full in The Verse Dieter Roth Missed.

    Fifty-Eight Verses the Story Needs

    A gospel is not a list of sayings. Its stories run in order, and a verse that is attested often makes no sense without the verse before it. Where that is so, I keep the verse as probable. Most of these verses have no witness at all; a few, like the healing on the sabbath in Luke 13, Tertullian reports only in general terms. Some examples:

    • Luke 9:37-39. Epiphanius quotes Luke 9:40 from Marcion’s text: “I begged your disciples to cast it out, but they weren’t able” (Evangelion 6:4 TT). That sentence needs a speaker and an “it.” The three verses before it supply both, so I keep them: “On the day Jesus came down from the mountain, a large crowd met him” (Evangelion 6:1 TT).
    • Luke 23:4-5. Pilate’s sending of Jesus to Herod (Luke 23:7) is attested. It assumes Pilate’s question in 23:6, which I also keep as probable, and that question assumes the charge in 23:5 and the verdict it answers in 23:4: “Pilate said to the chief priests and the crowd, ‘I find nothing wrong with this man’” (Evangelion 15:22 TT).
    • Luke 24:28-29. The meal at Emmaus is attested. The two verses that bring Jesus to the table are not, but the meal cannot happen without them: “But they urged him: ‘Stay with us. It’s nearly evening, and the day is almost over.’ So he went in to stay with them” (Evangelion 16:25 TT).
    • Luke 6:15. The list of the twelve is attested at its beginning and its end. The four names in the middle are needed to make twelve: “Matthew, and Thomas; James the son of Alphaeus, and Simon, known as the Zealot;” (Evangelion 3:19 TT).

    None of these is presented as attested. Each carries the label Probable, so a reader can see what rests on evidence and what rests on the shape of the story. Not every scholar agrees with each of them: BeDuhn leaves out six of these verses and Klinghardt three, and Harnack thought Luke 24:44 struck out. The Critical Edition gives my reasons where they differ.

    Seventy-Three Verses Judged Absent

    Silence can also be evidence, when a witness passes over a passage he would have been glad to use. I judge 73 of Roth’s unattested verses absent. For four of them (Luke 4:14-15 and 21:23-24) a Greek witness reports the absence directly. For the rest I rely on the witnesses’ silence as the editors weighed it, Harnack and Zahn above all.

    • Luke 7:29-35, the children in the market-place. Tertullian passes straight from 7:28 to the woman who anointed Jesus. Zahn says the way he skips ahead, “although the content of 29–35 would have furnished especially abundant occasion for polemic, establishes the absence.” Harnack agrees that the passage “appears to have been missing.”
    • Luke 8:49-56, the raising of Jairus’s daughter. Harnack: “no trace of it is attested.” Zahn finds no trace of it in Tertullian either.
    • Luke 23:35-42, the mockery and the two criminals. Harnack judges that 35-43 “seems to have been entirely missing.” For the next verse, 23:43, “Today you will be with me in paradise,” which Roth himself counts absent, the evidence is direct: Epiphanius reports that Marcion’s text did not have it.
    • Luke 4:17-22, the reading from Isaiah at Nazareth. Tertullian writes that at Nazareth “nothing new is noted as preached” (nihil novi notatur praedicasse, 4.8.2). Harnack: “M. has not stated the content of the sermon.” Zahn brackets the verses as absent.

    Harnack and Zahn do not always agree, and the Critical Edition shows where they part. On the drowning of the pigs (Luke 8:33-37), Harnack judges the story struck out because Tertullian says only “they obtained it,” while Zahn prints the verses as present. I follow Harnack and give both views. The full case is in The Pigs That Never Ran Off the Cliff.

    Two Hundred Twenty-Six Verses No One Can Decide

    For the rest, I agree with Roth and BeDuhn: the evidence is not there. These verses are not in the True Testament, but I do not claim they were not in Marcion’s Gospel. The Critical Edition labels each passage “absence not established” and prints its Greek in double brackets at the place where it stood, so that the silence is kept visible.

    Often the German editors said the same thing a century ago. On Luke 4:36-39, the healing of Peter’s mother-in-law, Harnack wrote: “nothing is known.” Zahn: “we know nothing at all.” Ritschl, in 1846, thought the story probably missing, because the healed woman serves “them” before any disciple has been called. On the walk to the Mount of Olives (Luke 22:39-40), Harnack: “unattested.” On the plotting of the Pharisees (Luke 11:53-54), both Harnack and Zahn: unattested.

    Where the evidence runs out, the Critical Edition says so. It does not fill the gap from Luke, as Hahn’s text often does, and it does not leave the reader to guess what “unattested” will mean in the end.

    BeDuhn’s 301

    BeDuhn’s 301 unattested verses come out the same way. The Critical Edition finds a witness for 6 of them, keeps 17 as probable, judges 61 absent (2 on a Greek witness’s report), and agrees that 217 cannot be decided.

    What This Shows

    Roth and BeDuhn were right to be careful. But a reader asking what Marcion’s Gospel contained deserves more than a third of Luke left blank. The fathers left evidence that reaches further than “unattested” allows: Tertullian reports scenes he does not quote, his silences say something where he had every reason to speak, and an attested verse often carries the verses it depends on. Harnack and Zahn used all of this a century ago.

    The Critical Edition uses it too, verse by verse, and labels the result. Of the 385 verses Roth leaves open, by our count, it decides 159 and leaves 226 open, saying so in each place. Every one of these verses has its entry in the Critical Edition of the Greek True Testament, and each can be set beside Roth, BeDuhn, Klinghardt, Harnack, Zahn and the others on the Compare Reconstructions page.