Are the Gospels reliable history?

The short version

1. We have what they wrote.

   About 5,700 Greek manuscripts; substantial copies within about 150 years; the variants that matter are known. That is the starting line, not a reason to believe.

   The evidence ↓
2. They were written within living memory.

   Paul's letters in the 50s, Mark about 65-75, all four Gospels by about 100. The case does not rest on who wrote them.

   The evidence ↓
3. The writers knew the place.

   The names fit first-century Palestinian Jewish naming, not Diaspora naming or novels. Small towns, pools and local Roman titles check out. That rules out distant invention, not Palestinian-rooted tradition.

   The evidence ↓
4. Outsiders confirm the outline.

   Tacitus, Josephus and Pliny confirm Jesus "called Christ", executed under Pilate and worshipped within about 80 years. Inscriptions and excavations confirm Pilate, very probably Caiaphas, and two pools in John.

   The evidence ↓
5. Serious sources, with real flaws.

   The accounts differ in detail, Luke makes two probable errors, and Matthew tells some details to fit Scripture. Overall the evidence is about 8 times likelier if these are early, rooted testimony than late legend (range 1.2 to 80), but barely favours them over the mainstream critical view: serious sources, nowhere near enough by themselves to carry a miracle claim.

   The analysis ↓

The question

Judged by the methods historians use on any ancient source, do the Gospels and Acts preserve substantially reliable early testimony about Jesus and the first Christians? The question is about their non-miraculous content: people, places, offices, the sequence and setting of events, the teaching. Whether the miracle reports are true is the next question, and this page does not answer it. It decides only whether those reports can be waved away as late legend.

We compare two hypotheses:

• H, reliable early testimony. The texts were written within living memory (roughly AD 50-100) by authors in touch with Palestinian Jewish tradition and eyewitness-era sources. Luke says as much of himself: "those who from the beginning were eyewitnesses and servants of the word delivered them to us" (Luke 1:2, WEB).¹ They were copied faithfully. The core facts are accurate, with the ancient biographer's normal freedom in order, compression, paraphrase and emphasis.
• L, late legendary development. The narratives are mainly the product of decades of community storytelling and theological invention, written far from the events in place, time or culture. L still allows a historical Jesus behind the stories. It predicts that the details (names, geography, offices, social texture) drift toward what a later, Diaspora or Gentile author would know.

Most critical scholars (Ehrman, Sanders and Dunn, for example) hold neither. They see a historical core with substantial memory distortion and theological shaping, in texts that are early-ish but not eyewitness accounts. Part of that shaping is telling details so that they fit Scripture; the messianic prophecy page calls the rate at which that happens f, and the same f applies here. The evidence below separates H and this middle position from L well. It separates H from the middle position only weakly (a ratio of about 1.3), and we say so wherever it matters.

Rules of evidence

We fixed these before scoring anything:

1. Text and truth are separate. Manuscripts tell us what was written, not whether it is true. Bart Ehrman's point, and we grant it: the manuscript evidence is a precondition and is not scored.
2. A confirmation must be hard to fake. It is a checkable detail confirmed by something outside the New Testament (an inscription, an excavation, a non-Christian text, a database of names). It counts only if a late or distant writer would plausibly get it wrong, and only if the author could not simply have copied it from a source we still have. Josephus is the main worry for Luke and Acts.
3. A miss is a miss. A detail contradicted by better evidence, or two accounts that cannot both be true on any reading an ancient reader would accept, counts against. If a harmonisation needs special pleading, we score it as a miss even though one exists.
4. The null space goes in. Every weighty alleged contradiction or error goes into the table, not just the easy ones.
5. Controls. Where we can, we compare the Gospels with texts we know are late or fictional (the second-century apocryphal gospels, historical novels, Talmudic legends) and with a text we know is historical (Josephus). A feature counts only if it separates the two groups.
6. Excluded: miracle claims (the resurrection page), the Shroud, disputed artefacts (the James ossuary, the Nazareth Inscription, the "Jesus boat" as Jesus' boat), and the infancy narratives, which we weigh separately and keep out of the total.
7. Overlap is discounted. Names, places and offices all reflect one thing, knowledge of Palestine before 70, so their combined weight is shrunk.
8. "Reliable" means the setting and the outline. Nothing scored here checks the details that only the Gospels report and that they present as fulfilling Scripture. Where such a detail is visibly shaped, we count it once, as a cost, in the errors line.

The evidence

1. Do we have what they wrote?

The official register of Greek New Testament manuscripts, kept at the Institute for New Testament Textual Research in Münster, put the number of surviving Greek manuscripts at about 5,700 in 2023.^{2, 3} Ehrman's own count agrees: "over 5700 Greek manuscripts discovered and catalogued".⁴ The often-quoted 24,000 or 25,000 adds translations into Latin, Syriac, Coptic and other languages, about 10,000 of them Latin.⁵ Most manuscripts are medieval, and only a few contain the whole New Testament.

The fair comparison is with other ancient works. The usual apologetics tables are out of date on the classical side, so we use Clay Jones's recount from the classical editions themselves.⁵

papyri       highest no.   141  removed    6  extant ~  135
majuscules   highest no.   326  removed   43  extant ~  283
minuscules   highest no.  3019  removed  159  extant ~ 2860
lectionaries highest no.  2555  removed  133  extant ~ 2422
TOTAL        highest nos. 6041  removed  341  extant ~ 5700

2. When were they written, and by whom?

Paul's undisputed letters (about 48-60) are the earliest Christian texts. They already confirm Jesus' crucifixion, James "the Lord's brother" (Galatians 1:19), Peter and the Twelve, the Last Supper tradition (1 Corinthians 11:23-25) and the early creed of 1 Corinthians 15:3-7. On the majority critical dates, Mark follows about 65-75, Matthew and Luke about 75-90, Acts about 80-95 and John about 90-100. Ehrman's own dates are close to these.¹⁷

On the majority dates, all four Gospels fall about 35-70 years after the events (the early-date case makes the gap shorter), while eyewitnesses and those who heard them were still alive. That is far closer than the main sources for Alexander the Great (Arrian and Plutarch, more than 400 years later) and comparable to Tacitus on Tiberius (about 80 years). The window for legend is short but not zero. A. N. Sherwin-White's often-quoted claim, that two generations were too short for myth to displace the historical core, is an argument, not a law.¹⁸

3. The names test

Names are hard to fake. In any population a few names are very common and most are rare, and the pattern differs from place to place. Tal Ilan's Lexicon of Jewish Names in Late Antiquity catalogues about 3,000 named Jews of Palestine from 330 BC to AD 200, drawn from inscriptions, ossuaries, papyri, Josephus and rabbinic texts.²⁶ Richard Bauckham compared the named Palestinian Jews of the Gospels and Acts with it:^{23, 27}

In 2024 Luuk van de Weghe and Jason Wilson ran formal goodness-of-fit tests on the full name lists.²⁸ (A p-value here asks: if these names were drawn from the reference population, how often would chance alone produce a mismatch at least this large? A large value means no detectable mismatch; a tiny one means chance cannot account for the mismatch, so the names do not follow that pattern.) They restricted Ilan's data to 4 BC-AD 73 and removed the Gospel and Josephus names from the reference, so the samples are independent.

Their conclusion: names in the Gospels and Acts "fit into their historical context at least as good as those in the works of Josephus".²⁸ Our own recomputation from Bauckham's published percentages lands in the same place (below). The Gospels also do what a real population forces: the commonest names get qualifiers ("Simon Peter", "Simon the Zealot", "Simon of Cyrene", "Mary Magdalene").²⁹ The late apocryphal gospels lack this texture. The Gospel of Thomas names seven people and hardly any places; the Gospel of Judas has only two recognisable Palestinian Jewish names.²⁹

1. Implied Gospels-Acts male sample: n=77 (Simon/Joseph=14, top-9=31, attested-once=3)
2. Exact binomial tests, Gospels-Acts vs Ilan (two-sided), 95% Clopper-Pearson intervals:
   Simon or Joseph               GA 14/77 = 18.2%  [95% CI 10.3-28.6%]  Ilan 15.6%  p = 0.529
   one of the 9 commonest names  GA 31/77 = 40.3%  [95% CI 29.2-52.1%]  Ilan 41.5%  p = 0.908
   name attested only once       GA  3/77 =  3.9%  [95% CI  0.8-11.0%]  Ilan  7.9%  p = 0.286
   women: Mary or Salome         GA  7/18 = 38.9%  [95% CI 17.3-64.3%]  Ilan 28.6%  p = 0.433
3. Three-cell multinomial (Simon/Joseph | other 7 of top 9 | other), counts = [14, 17, 46]
   log10 LR (Palestinian popularity : uniform inventor over 457 names) = 32.3
   Mixture "fraction f of names from real Palestinian popularity, the rest uniform":
   posterior mean f = 0.88; one-sided 95% lower bound f >= 0.72
   LR(f=1 : f=0.90) = 1.11   LR(f=1 : f=0.75) = 3.58   LR(f=1 : f=0.50) = 840   LR(f=1 : f=0.25) = 6e8

4. Places and offices

A writer far away gets the famous places right and the small ones vague. Local Roman titles, which changed by province and decade, are a sharper test still. The Gospels and Acts pass several checks that critics once thought they failed:

• The pool of Siloam (John 9:7) was found in 2004: a stepped, plastered Second Temple pool.³⁴ The pool of Bethesda, "by the sheep gate ... having five porches" (John 5:2, WEB), turned out to be twin pools north of the Temple; four sides and a dividing wall give five colonnades.³⁵ Both are in John, long thought the least historical Gospel.
• Politarchs. Acts 17:6-8 calls the magistrates of Thessalonica politarchai. More than 60 inscriptions use the title, about three-quarters of them from Macedonia and about half from Thessalonica itself.³⁶
• Gallio was "proconsul of Achaia" (Acts 18:12, WEB). A letter of the emperor Claudius inscribed at Delphi dates his term to about 51-52, which anchors the chronology of Paul's travels.³⁷
• Pilate's title. A stone found at Caesarea in 1961 calls him prefect of Judaea. Tacitus, writing later, anachronistically calls him procurator.^{38, 39}

5. What outsiders said

Three writers with no stake in the Christian story mention Jesus or his followers within a century:

"Christus, the founder of the name, had undergone the death penalty in the reign of Tiberius, by sentence of the procurator Pontius Pilatus."Tacitus, Annals 15.44, c. 116 (Loeb translation)³⁹

"... the brother of Jesus, who was called Christ, whose name was James ..."Josephus, Antiquities 20.200, c. 93-94, on the high priest Ananus' trial of James in 62 (Whiston translation)⁴¹

Christians "met on a stated day before it was light, and addressed a form of prayer to Christ, as to a divinity".Pliny the Younger, Letters 10.96, c. 111-112, from Bithynia (Melmoth translation)⁴⁷

Together they confirm that Jesus existed, was called Christ and was executed by Pilate under Tiberius; that his brother James was put to death in 62; and that a movement worshipping him reached Rome by the 60s and Asia Minor by about 110. They confirm nothing distinctive about his deeds, and nothing about the resurrection. Because L also keeps a historical core, this line separates the hypotheses only weakly.

6. Undesigned coincidences and embarrassment

An undesigned coincidence is when one account leaves a small puzzle that another answers in passing, without seeming to intend it.^{56, 57} The best example: at the feeding of the five thousand, John has Jesus turn to Philip, of all the disciples, and ask "Where are we to buy bread, that these may eat?" (John 6:5, WEB). John does not say where they were. Luke does: "a city called Bethsaida" (Luke 9:10, WEB). And John mentions elsewhere, for other reasons, that "Philip was from Bethsaida" (John 1:44, WEB). Neither writer seems to be building the link.

A second: in Luke, Pilate hears that Jesus claims to be "Christ, a king", asks him one question, and announces "I find no basis for a charge against this man" (Luke 23:2-4, WEB). Why? John supplies the private exchange: "My Kingdom is not of this world" (John 18:36, WEB).

The critics' replies are serious.⁵⁸ The Synoptics copy each other and John probably knew Mark, so only links between independent strands count. Nobody has measured how often novels or legends produce coincidences of this quality, so there is no base rate. And only the fits get counted. We keep a handful of strong cases, drop the weak ones, and give this line a floor of 1, meaning no evidence at all on the critics' best reading.

Embarrassment. The texts keep details that cut against their own interest: Jesus receiving John's "baptism of repentance" (Mark 1:4, WEB), which Matthew visibly struggles with (Matthew 3:14-15); his own people saying "He is insane" (Mark 3:21, WEB); "Why do you call me good?" (Mark 10:18, WEB); his ignorance of the day (Mark 13:32); Peter's denial and the disciples' flight; Jesus rounding on Peter, the leading apostle, with "Get behind me, Satan! You are a stumbling block to me" (Matthew 16:23, WEB; Mark 8:33), kept by a tradition that revered Peter; women as the first witnesses.⁵² The criterion is contested (what embarrassed whom?), and Josephus' rule that women's testimony was not accepted was not absolute.⁴¹ A modest factor.

7. Where the accounts differ, and where they are wrong

Richard Burridge showed that the Gospels share the marks of Greco-Roman biography (bios), now the majority view.⁶¹ Ancient biographers aimed at truth about a person but allowed topical order, speeches in their own words, compression, and a focus on the subject's death. Licona catalogues the same devices in Plutarch's Lives that explain many Gospel differences.⁶⁰ The fair summary: reliable by ancient standards, not reliable as a modern transcript.

Memory science points the same way. Ehrman argues that memories, including a community's, are reconstructive and distort quickly.⁶² Bauckham argues that named eyewitnesses remained guarantors of a controlled tradition.²³ Both are partly right. Memory tends to keep the gist, the central and the emotionally charged, and to lose peripheral detail, wording and order. That is the pattern of the Gospel differences: agreement on the core, variation in order and detail.

Two problems are real, and we score them as probable errors:

• Quirinius. Luke dates Jesus' birth to "the first enrollment made when Quirinius was governor of Syria" (Luke 2:2, WEB), but both Luke (1:5) and Matthew put the birth under Herod, who died in 4 BC, and Josephus dates Quirinius' census of Judea to AD 6.⁴¹ The proposed defences are strained. Probably a mistake or a conflation by Luke.
• Theudas. In Acts 5:36-37, Gamaliel, speaking about 33-35, cites Theudas and then "after this man, Judas of Galilee" (AD 6). Josephus puts Theudas about 44-46.⁴¹ Either there was an earlier, unknown Theudas, or Luke slipped, perhaps misreading Josephus.²²

Add a probable slip in Mark (Abiathar, below). These cost the Gospels some credit for precision in detail, not for the core.

8. Told to fit Scripture?

Some details were. Zechariah 9:9's poetic parallel, quoted by Matthew as "riding on a donkey, on a colt, the foal of a donkey" (Matthew 21:5, WEB), becomes two animals in his story: "you will find a donkey tied, and a colt with her", "and he sat on them" (Matthew 21:2, 7, WEB). Mark, Luke and John have one animal. Matthew also says Jesus lived in Nazareth "that it might be fulfilled which was spoken through the prophets that he will be called a Nazarene" (Matthew 2:23, WEB), and no Old Testament text says that. Bethlehem has an obvious motive in Micah 5:2 (above).

Nothing else on this page tests details like these. Names, places, offices and outside writers confirm the setting and the outline, and a writer who knew Palestine well could still tell a detail to fit a text. So this page gives no reason to think such shaping was rare. On the messianic prophecy page the whole result turns on that rate, and Matthew's visible cases put it at least about 1 in 10 for claims nobody could check; this page is consistent with that and does not lower it. What the shaping does not reach is the outline that Paul and outside writers confirm: the crucifixion under Pilate, James, the early worship.

The analysis

The argument runs on the general rules of source criticism used for any document. A text is a reliable source to the degree that it was written close to the events by people with access to informants, its checkable details are right when checked independently, its accounts agree on the core, and its text was transmitted accurately. The Gospels and Acts score well on all four, with real faults in detail. By the same canons used for Tacitus or Josephus, they are substantially reliable sources for the non-miraculous core of Jesus' life and the early movement.

To put a number on it we use a likelihood ratio: how much more probable the evidence is if H is true than if L is true. 1 means the evidence does not discriminate; 10 means it is ten times more expected under H. Only the names line has a statistical test behind it. The rest are judgments with explicit ranges, listed so you can change them.

dependence factor d = 0.50:  LR low 1.16   central  5.6   high   40
dependence factor d = 0.60:  LR low 1.19   central  8.0   high   84   <- reported
dependence factor d = 0.75:  LR low 1.24   central 13.4   high  253
dependence factor d = 1.00:  LR low 1.34   central 31.8   high 1600

Sensitivity (d = 0.6):
  drop the names line entirely                 -> central 4.6
  drop undesigned coincidences + embarrassment -> central 5.6
  errors line at its old value (0.6)           -> central 8.9
  errors line harsher (0.2)                    -> central 4.6
  a reader outside our ranges: names, places 1.5;
    lines 3-7 at 1; errors 0.3                 -> 0.79

Objections

What this shows, and what it doesn't

What we don't claim

• "24,000 (or 25,000) manuscripts." The Greek count is about 5,700; the rest are translations, and those counts are soft.
• "Homer 643, Caesar 10, Plato 7." Out of date. The classical editions give 1,757+, 251 and 210. The corrected numbers shrink the New Testament's lead but do not erase it.
• "P52 dates to AD 125," or "fragments within 25-50 years." The range is about 125-175, possibly later.
• "P66 and P75 date to about 200." Contested; they may be 4th-century.
• "First-century Mark"; "7Q5 is Mark"; "the Magdalen papyrus (P64) is first-century." False or fringe.
• "The New Testament is 99.5% accurate"; "no variant affects any doctrine." The first has no traceable derivation; the second overstates. No central doctrine rests only on a disputed reading.
• "More copies prove the Gospels true." Transmission is not truth.
• "The Synoptics were certainly written before 70." A minority view.
• "Luke is accurate in every detail"; "no archaeological discovery has ever contradicted the New Testament." Quirinius and Theudas.
• "Hemer verified 84 facts." His lists are details consistent with the setting, of varying force, not 84 independent proofs.
• "Ramsay called Luke a historian of the first rank." We have not verified the wording from Ramsay's text, so we don't use it.
• The full Testimonium Flavianum as Josephus' words. Only Antiquities 20.200 and, at most, a labelled reconstructed core of 18.63-64.
• Suetonius' "Chrestus" as certainly Jesus; Mara bar Serapion as a certain first-century reference; b. Sanhedrin 43a as an independent first-century witness; Thallus and Phlegon on the darkness. The last two survive only through a 3rd-century writer, Julius Africanus.
• "Women's testimony was worthless in court." It carried little weight; the rule was not absolute.
• The Nazareth Inscription as evidence for the empty tomb; the James ossuary; the "Jesus boat" as Jesus' boat; the 2018 Herodium ring as Pilate's own. Unknown provenance, disputed, or no link.
• "The names prove eyewitness testimony," or any ratio like 10³⁰ from the names. They exclude distant invention, not Palestinian-rooted tradition.
• Bethlehem as established history, or as a fulfilled prophecy confirmed by the reliability case. Possible, not established.
• "Nazareth didn't exist." The skeptic's counter-myth, refuted by excavation.
• "Mark 16:9-20 and John 7:53-8:11 are original," or that their absence shows the text is corrupt. They are known later additions, flagged in every modern Bible. So is the closing line of the Lord's Prayer.
• "The Gospels are reliable history, so the prophecies were fulfilled just as they report." Reliability here covers the setting and the outline. Matthew tells some fulfilment details to fit the text, and this page gives no reason to think that was rare.
• "Our low end is what a skeptic would say." It is the floor of our own ranges. A reader outside them reaches about 0.8.
• "The Gospels are eyewitness transcripts," or the wording of John's "I am" discourses as Jesus' own words, established by this case. Ancient histories wrote speeches to the general sense; John's discourses are the least secure layer of the sayings.
• The parables, the "I am" sayings, the Lord's Prayer or the three temptations as encoded physics, or a "complete set", whose fit proves a superhuman source. A framework built from these texts will fit them; the fit is circular, not confirmation. "Complete" holds only because the categories were chosen to cover everything, and the temptations even come in two orders.
• A physical mechanism for Jesus' miracles or for faith moving mountains (faith as a quantum operator, matter held in place by "collective observation"). Invented terms resting on a minority reading of quantum mechanics, with no test; and miracles are not this page's question.

Sources

1. Bible quotations: the World English Bible (public domain), ebible.org/eng-web. Its footnote to Matthew 6:13 reads "NU omits" the closing doxology (NU: the Nestle-Aland/UBS critical text). Checked: read.
2. Leggett, K. & G. Paulson (2023). "How many Greek New Testament manuscripts are there, really? The latest numbers." INTF blog, Münster. ntvmr.uni-muenster.de. The page did not load for us; figures as tabulated by Gurry (next source).
3. Gurry, P. J. (2023). "How many Greek NT manuscripts are there?" (25 Oct) and "Explaining a Discrepancy in the Number of Greek NT Manuscripts" (2 Nov). Evangelical Textual Criticism blog. 25 Oct post; INTF label. Checked: read in full.
4. Ehrman, B. D. (2017). "What we now know about the manuscripts of the New Testament." The Bart Ehrman Blog, 14 Feb. ehrmanblog.org. Checked: read in full.
5. Jones, C. (2012). "The Bibliographical Test Updated." Christian Research Journal 35(3). equip.org. Compiles the classical counts from M. L. West (2001), T. W. Allen (1931), Brumbaugh & Wells (1968), V. Brown (1979), L. D. Reynolds (1983) and D. Potter (1999). Checked: read in full. (The classical editions themselves: not yet checked by us.)
6. On Homeric papyri in the Leuven Database of Ancient Books (about 2,000). Pylon (Heidelberg), 2025. journals.ub.uni-heidelberg.de. Not yet checked by us; reported in secondary sources.
7. Nongbri, B. (2005). "The Use and Abuse of P52: Papyrological Pitfalls in the Dating of the Fourth Gospel." Harvard Theological Review 98: 23-48. Not yet checked by us; reported in secondary sources.
8. Nongbri, B. (2020). "Palaeography, Precision and Publicity: Further Thoughts on P.Ryl. III.457 (P52)." New Testament Studies 66(4): 471-499. cambridge.org. Checked: abstract only.
9. Orsini, P. & W. Clarysse (2012). "Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography." Ephemerides Theologicae Lovanienses 88: 443-474. Not yet checked by us; reported in secondary sources.
10. Colomo, D. & D. Obbink (eds.) (2018). P.Oxy. LXXXIII 5345 (P137). The Oxyrhynchus Papyri 83. Oxford portal. Not yet checked by us; reported in secondary sources.
11. Nongbri, B. (2014). "The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P.Bodmer II (P66)." Museum Helveticum 71: 1-35. Not yet checked by us; reported in secondary sources.
12. Nongbri, B. (2016). "Reconsidering the Place of Papyrus Bodmer XIV-XV (P75) in the Textual Criticism of the New Testament." Journal of Biblical Literature 135: 405-437. Checked: abstract only (via the Macquarie University repository record).
13. Ehrman, B. D. (2005). Misquoting Jesus: The Story Behind Who Changed the Bible and Why. HarperSanFrancisco. Not yet checked by us; reported in secondary sources.
14. Gurry, P. J. (2016). "The Number of Variants in the Greek New Testament: A Proposed Estimate." New Testament Studies 62: 97-121. repository.cam.ac.uk. Checked: abstract only. The 500,000 figure as reported by D. B. Wallace, danielbwallace.com.
15. Lanier, G. R. (2018). "Quantifying New Testament Textual Variants: Key Witnesses in Acts and the Catholic Letters." New Testament Studies 64. cambridge.org. Checked: abstract only.
16. Wallace, D. B., interviewed by J. Taylor, part 4. The Gospel Coalition. thegospelcoalition.org. Checked: read in full.
17. Ehrman, B. D. (2009). Jesus, Interrupted; (2011). Forged: Writing in the Name of God. HarperOne. Not yet checked by us; reported in secondary sources.
18. Sherwin-White, A. N. (1963). Roman Society and Roman Law in the New Testament. Oxford: Clarendon. Not yet checked by us; reported in secondary sources.
19. Crossley, J. (2004). The Date of Mark's Gospel. T&T Clark. Not yet checked by us; reported in secondary sources.
20. Robinson, J. A. T. (1976). Redating the New Testament. SCM. Not yet checked by us; reported in secondary sources.
21. Seccombe, D. (2020). On the date of Acts. Tyndale Bulletin. tyndalebulletin.org. Not yet checked by us; reported in secondary sources.
22. Pervo, R. I. (2006). Dating Acts: Between the Evangelists and the Apologists. Polebridge. westarinstitute.org. Not yet checked by us; reported in secondary sources.
23. Bauckham, R. (2006; 2nd ed. 2017). Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony. Eerdmans. Not yet checked by us; reported in secondary sources.
24. Hengel, M. (1985). Studies in the Gospel of Mark. SCM. Not yet checked by us; reported in secondary sources.
25. Eusebius, Ecclesiastical History 3.39.15 (quoting Papias on Mark). Not yet checked by us; reported in secondary sources.
26. Ilan, T. (2002). Lexicon of Jewish Names in Late Antiquity, Part I: Palestine 330 BCE-200 CE; Part III: The Western Diaspora (2008). Mohr Siebeck. Machine-readable data: eRabbinica (GitHub). Not yet checked by us; reported in secondary sources.
27. Bauckham's percentages as reported in: "Are the Gospels based on eyewitness testimony? The test of personal names." Christian Apologetics Alliance, 2012. christianapologeticsalliance.com. Checked: read. It gives the Simon/Joseph, top-nine and Mary/Salome figures; the "attested only once" pair (7.9% and 3.9%) is not in it and still needs checking against Bauckham's book.
28. van de Weghe, L. & J. Wilson (2024). "Why Name Popularity is a Good Test of Historicity: A Goodness-of-fit Test Analysis on Names in the Gospels and Acts." arXiv:2403.14883. arxiv.org/abs/2403.14883. Checked: read in full.
29. Williams, P. J. (2018). Can We Trust the Gospels? Crossway. Reviews: The Gospel Coalition; Theopolis. Not yet checked by us; reported in secondary sources.
30. Gregor, K. & B. S. Blais (2023). "Is Name Popularity a Good Test of Historicity? A Statistical Evaluation of Richard Bauckham's Onomastic Argument." Journal for the Study of the Historical Jesus 21(3): 171-202. digitalcommons.bryant.edu. Checked: abstract only.
31. Gregor, K. & B. S. Blais (2024). "Re-Examining Richard Bauckham's Argument from Name Popularity in Light of Apocryphal and Talmudic Evidence." Journal for the Study of the New Testament 47(2): 271-296. Not yet checked by us; reported in secondary sources.
32. Wilson, J. & L. van de Weghe (2025). "Re-Examining the Statistical Methodology and Onomastic Claims of Gregor and Blais' Argument from Name Popularity." arXiv:2508.21150. arxiv.org/abs/2508.21150. Checked: abstract only.
33. van de Weghe, L. (2023). "Name Recall in the Synoptic Gospels." New Testament Studies 69(1): 95-109. Not yet checked by us; reported in secondary sources.
34. Reich, R. & E. Shukron, the Siloam pool (excavated 2004). Reports: Biblical Archaeology Society; NBC News, 2005. Not yet checked by us; reported in secondary sources.
35. von Wahlde, U. C. (2006). "The Pool(s) of Bethesda and the Healing in John 5." In J. H. Charlesworth (ed.), Jesus and Archaeology. Eerdmans. Not yet checked by us; reported in secondary sources.
36. Burton, E. D. (1898). "The Politarchs." American Journal of Theology 2; G. H. R. Horsley in New Documents Illustrating Early Christianity; Fant, C. E. & M. G. Reddish (2008). Lost Treasures of the Bible. Eerdmans. Not yet checked by us; reported in secondary sources.
37. The Gallio inscription, Delphi (Claudius' letter), Sylloge Inscriptionum Graecarum³ 801. Not yet checked by us; reported in secondary sources.
38. Frova, A. (1961), publication of the Pilate inscription from Caesarea. Not yet checked by us; reported in secondary sources.
39. Tacitus, Annals 15.44. Loeb Classical Library, vol. V (1937), via LacusCurtius. penelope.uchicago.edu. Checked: read in full.
40. Greenhut, Z. and R. Reich (1992), reports on the Caiaphas family tomb and ossuary, Jerusalem. Not yet checked by us; reported in secondary sources.
41. Josephus, Antiquities of the Jews (4.219; 18.1-2; 18.3.3; 20.9.1 = 20.200; 20.97-98), tr. W. Whiston. Project Gutenberg #2848. gutenberg.org. Checked: 18.3.3 and 20.9.1 read; the other passages as cited by the scholarly literature.
42. Smith, J. (of Jordanhill) (1848). The Voyage and Shipwreck of St. Paul. Not yet checked by us; reported in secondary sources.
43. Hemer, C. J. (1989). The Book of Acts in the Setting of Hellenistic History, ed. C. H. Gempf. Mohr Siebeck. Not yet checked by us; reported in secondary sources.
44. Corbo, V. & S. Loffreda, Capernaum excavation reports. Not yet checked by us; reported in secondary sources.
45. Dark, K. R. (2020). Roman-Period and Byzantine Nazareth and its Hinterland and The Sisters of Nazareth Convent. Routledge. University of Reading press release (2015): reading.ac.uk. Not yet checked by us; reported in secondary sources.
46. Zias, J. & E. Sekeles (1985). "The Crucified Man from Giv'at ha-Mivtar: A Reappraisal." Israel Exploration Journal 35: 22-27. Not yet checked by us; reported in secondary sources.
47. Pliny the Younger, Letters 10.96, tr. W. Melmoth. Project Gutenberg #2811. gutenberg.org. Checked: read in full.
48. Carrier, R. (2014). "The Prospect of a Christian Interpolation in Tacitus, Annals 15.44." Vigiliae Christianae 68. Not yet checked by us; reported in secondary sources.
49. Suetonius, Claudius 25.4 and Nero 16.2; Lucian, The Passing of Peregrinus 11-13; Babylonian Talmud, Sanhedrin 43a. Not yet checked by us; reported in secondary sources.
50. Van Voorst, R. E. (2000). Jesus Outside the New Testament. Eerdmans. Not yet checked by us; reported in secondary sources.
51. Merz, A. & T. Tieleman (2008). "The Letter of Mara bar Sarapion." In Empsychoi Logoi. Brill, 107-133. Not yet checked by us; reported in secondary sources.
52. Meier, J. P. (1991). A Marginal Jew: Rethinking the Historical Jesus, vol. 1. Doubleday. Not yet checked by us; reported in secondary sources.
53. Whealey, A. (2003). Josephus on Jesus. Peter Lang; (2008). "The Testimonium Flavianum in Syriac and Arabic." New Testament Studies 54. Not yet checked by us; reported in secondary sources.
54. Olson, K. A. (1999). "Eusebius and the Testimonium Flavianum." Catholic Biblical Quarterly 61. Not yet checked by us; reported in secondary sources.
55. Schmidt, T. C. (2025). Josephus and Jesus: New Evidence for the One Called Christ. Oxford University Press (open access). Reviews: Roger Pearse; History for Atheists. Not yet checked by us; reported in secondary sources.
56. Blunt, J. J. (1847). Undesigned Coincidences in the Writings both of the Old and New Testament, an Argument of their Veracity. Not yet checked by us; reported in secondary sources.
57. McGrew, L. (2017). Hidden in Plain View: Undesigned Coincidences in the Gospels and Acts. DeWard. Not yet checked by us; reported in secondary sources.
58. Blais, B. (2022). "Why undesigned coincidences are not compelling to me." Blog, 11 Jul. bblais.github.io. Checked: read in full.
59. McGrew, L. (2019). The Mirror or the Mask: Liberating the Gospels from Literary Devices. DeWard. Not yet checked by us; reported in secondary sources.
60. Licona, M. R. (2017). Why Are There Differences in the Gospels? What We Can Learn from Ancient Biography. Oxford University Press. Not yet checked by us; reported in secondary sources.
61. Burridge, R. A. (1992; 2nd ed. 2004; 3rd ed. 2018). What Are the Gospels? A Comparison with Graeco-Roman Biography. Cambridge University Press; Eerdmans; Baylor. Not yet checked by us; reported in secondary sources.
62. Ehrman, B. D. (2016). Jesus Before the Gospels. HarperOne. Not yet checked by us; reported in secondary sources.
63. Thucydides, History of the Peloponnesian War 1.22, tr. R. Crawley. Project Gutenberg #7142. gutenberg.org. Checked: passage read.
64. Justin Martyr, Dialogue with Trypho 78; Origen, Against Celsus 1.51; Jerome, Letter 58. Not yet checked by us; reported in secondary sources.