From Claude.ai on 6/21/2026.
The Theudas–Judas chronological problem in Acts 5:36–37 is widely regarded among critical scholars as one of the most damaging historical errors in Luke-Acts, precisely because it is not a minor slip but a structural anachronism embedded in a speech Luke attributes to a historical figure.
The Text
Acts 5:36–37 (NRSV):
“For some time ago Theudas rose up, claiming to be somebody, and a number of men, about four hundred, joined him; but he was killed, and all who followed him were dispersed and disappeared. After him Judas the Galilean rose up at the time of the census and got people to follow him; he also perished, and all who followed him were scattered.”
The Josephan Evidence
The problem is triangulated through two passages in Josephus. In Antiquities 20.97–99, Josephus describes Theudas’s revolt during the procuratorship of Fadus (44–46 CE), in which Theudas led followers to the Jordan promising to part its waters before being captured and beheaded by Roman cavalry. In Antiquities 18.4–10, Josephus places Judas the Galilean’s revolt precisely at the census of Quirinius in 6 CE — roughly four decades before Theudas.
Luke’s Gamaliel, speaking in a Jerusalem council scene set in the early 30s CE, reverses this order entirely: he invokes Theudas as a prior example, then Judas as a subsequent one. This is chronologically impossible on any reading of Josephus.
Scholarly Analysis
Steve Mason, in Josephus and the New Testament (2003), develops what remains the most sustained critical treatment of Luke’s literary relationship to Josephus. Mason argues that the inversion is best explained not by independent oral tradition but by Luke’s misreading of Josephus. In Antiquities 20, Josephus mentions Theudas, and then — in a parenthetical observation — refers back to Judas the Galilean and his sons (Antiquities 20.102), noting that the sons were crucified under Felix. Mason contends that Luke, reading this passage, encountered Theudas first and Judas second in the text of Josephus, and reproduced that literary sequence without grasping that Josephus’s mention of Judas was a retrospective reference to an earlier event. The error, on Mason’s reading, is one of a reader failing to navigate Josephus’s non-linear narrative structure.
Richard Carrier, in On the Historicity of Jesus (2014) and in earlier journal work, treats the Theudas–Judas inversion as among the strongest pieces of evidence for Lukan dependence on Josephus, and therefore as strong evidence against Acts’ compositional date being placed in the 60s CE. Carrier is characteristically direct: if Luke had access to independent, reliable oral or documentary tradition about these revolts, the probability of inverting their sequence approaches zero. The inversion makes sense only if Luke derived the names from a source — Josephus — where the sequence was textually Theudas-then-Judas, even though historically it was the reverse.
Bart Ehrman, while not developing the argument at the length Mason and Carrier do, acknowledges in The New Testament: A Historical Introduction to the Early Christian Writings that Acts contains demonstrable historical errors when checked against Josephus, and that the Theudas passage is a primary exhibit. Ehrman locates this within his broader skepticism about the historical reliability of Acts as a source for early Christian history.
Richard Pervo, in Dating Acts (2006) — arguably the most comprehensive critical monograph on the question — uses the Theudas problem as a pillar of his argument that Acts was composed no earlier than the early second century, after Josephus’s Antiquities (completed c. 93–94 CE) had begun circulating. Pervo builds a cumulative case from multiple Josephan parallels, but treats the Theudas–Judas inversion as particularly probative because it is not merely a shared detail but a shared error of sequencing — a type of mistake that strongly implies a single defective source rather than convergent independent traditions.
F.F. Bruce, representing the conservative response, argued in his commentary on Acts that Luke may have drawn on a separate tradition in which a different Theudas — unknown to Josephus — is referenced, and that the coincidence of names is incidental. This harmonizing move is sometimes called the “two Theudas” hypothesis.
Joseph A. Fitzmyer, in his Anchor Bible commentary on Acts (1998), takes the problem seriously without fully embracing the Josephan dependence hypothesis. He acknowledges that no satisfactory solution has been found, and that the error, if it is an error, points toward a late date of composition or significant source confusion, but he resists Pervo’s most aggressive conclusions.
Ernst Haenchen, in his landmark critical commentary The Acts of the Apostles (ET 1971), was earlier and blunter: he concluded that Luke simply made a mistake, that the speech of Gamaliel is a Lukan composition rather than a historical transcript, and that the anachronism reveals Luke’s hand inserting illustrative historical examples he did not fully control.
Why the “Two Theudas” Harmonization Fails
The F.F. Bruce escape route — positing a prior Theudas unknown to Josephus — is generally regarded by critical scholars as ad hoc for several reasons. First, Josephus is notably thorough in cataloguing Jewish messianic and revolutionary movements of the period, making it improbable that a movement large enough for Gamaliel to invoke as a cautionary example (four hundred followers) would leave no trace. Second, the structural parallel between Acts 5 and Josephus Antiquities 20 is too precise: both name Theudas, both give roughly the same follower count, both describe dispersal after the leader’s death. The probability that two independent Theudas movements of similar size, with the same leader name and the same outcome, occurred in the same region strains credulity beyond what methodological charity can sustain.
The Broader Significance
What makes this problem especially significant for the question of Acts’ date and sources is the type of error it represents. A random factual error — a wrong place name, a confused title — might plausibly arise from oral tradition or imperfect memory. But a sequential error, in which two events are placed in the wrong chronological order, and in which that wrong order exactly matches the textual order in which they appear in Josephus, is the kind of error that points specifically to literary dependence on a written source. Mason’s reconstruction — Luke reading Josephus and failing to recognize a retrospective reference — is, as Pervo notes, the most economical explanation that accounts simultaneously for the shared details, the shared error, and the directionality of the mistake.
