Do the Gospel Names Fit First-Century Palestine? The Onomastic Evidence

Imagine you are handed an anonymous short story and asked one question: where and when was the author actually from? You could try to catch them on geography or customs—but a clever forger can study a map. There is, however, one detail almost impossible to fake from the outside, because the writer is not even thinking about it: the names. Pick up a British novel set in 1950, and the boys are Brian and Keith and Derek; set it in 2025 and they are Oliver and Noah and Muhammad. No author sat down to research that. The names simply leaked out, carrying the unconscious watermark of a particular time and place. Get them subtly wrong—name a 1950 schoolboy “Jayden”—and a native reader knows at once that the writer was not there.

Now apply that test to the Gospels and Acts. They are populated by dozens of named people: fishermen, tax collectors, women at a tomb, a man who lent his grave. If these books were composed late and far away—Greek legends spun in Rome or Alexandria two or three generations after the fact, as much skeptical scholarship has long maintained—then the names should carry the watermark of that world, not the watermark of early-first-century Jewish Palestine. The remarkable claim of this essay’s argument is that they carry the Palestinian watermark with surprising statistical fidelity. The names are right—and right in a way that is extraordinarily hard to counterfeit.

This is a quieter argument than the case for the resurrection, and a narrower one. It overlaps with, but is not the same as, the fuller case that the Gospels rest on eyewitness testimony; here the focus is on the name data itself as a marker of provenance—of when and where these traditions were really formed. Used honestly, it is a piece of corroboration, not a proof. But it is a surprisingly stubborn piece.

Stating the Argument Carefully

The discipline at work is onomastics, the study of personal names and their distribution. Its founding observation is simple and uncontroversial: the popularity of a given name is sharply local and time-bound. Which names are common, and in what proportions, varies by region, by religious community, and by generation. This is not a contested theological premise; it is an everyday fact any historian of antiquity will grant.

From there the argument runs:

Premise 1: The relative frequency of personal names in a population is local and time-bound, and could not be reproduced by guesswork by an outsider or a much later author. Premise 2: The names in the Gospels and Acts match—in their specific names, their relative frequencies, and their disambiguation habits—the independently attested profile of Jewish Palestine in roughly 330 BC–AD 200, and do not match the different profile of the Jewish diaspora. Conclusion: Therefore the Gospel traditions are best explained as rooted in early Palestinian sources close to the events, rather than as late, geographically distant compositions.

Notice what this is and is not. It is an inference to the best explanation built on a statistical fact, not a deductive proof. Premise 1 is essentially conceded by everyone. The real work—and the real debate—lies in Premise 2 (does the data genuinely match, and how tightly?) and in how far that match licenses the conclusion. We will give both their due.

How the Argument Developed

The argument has a clear pedigree. Its raw material is the labor of the Israeli historian Tal Ilan, whose Lexicon of Jewish Names in Late Antiquity, Part I: Palestine 330 BCE–200 CE (Tübingen: Mohr Siebeck, 2002) catalogued every attested Jewish person of that region and era—drawn from literary texts, Josephus, ossuary inscriptions, and documentary papyri—and tabulated the frequency of their names.¹ Ilan was not doing apologetics; she was building a reference work. But she handed later scholars a hard, independent baseline against which any text claiming to depict that world could be tested.

The person who ran that test on the Gospels was the British New Testament scholar Richard Bauckham, in chapters 3 and 4 of Jesus and the Eyewitnesses: The Gospels as Eyewitness Testimony (Grand Rapids: Eerdmans, 2006).² Bauckham took Ilan’s tables and laid the Gospel names beside them. The argument had antecedents—earlier scholars had noticed the “Jewishness” of Gospel names—but Bauckham made it quantitative, and it became the most-discussed plank of a widely reviewed book. It has since been popularized by the Cambridge scholar Peter J. Williams in Can We Trust the Gospels? (2018), who folds it into a broader battery of “local knowledge” tests.³

Premise 1: Why Name Frequency Is a Watermark

It is worth dwelling on why name distribution is so hard to fake, because the strength of the whole argument depends on it. Three features combine.

First, names cluster by community and era in ways nobody plans. The fashion that makes “Simon” ubiquitous in one generation and rare in another is an emergent social fact, not a rule anyone consciously follows. You cannot deduce it; you can only absorb it by living inside the population.

Second, the diaspora pattern differed sharply from the Palestinian one. Among Egyptian Jews, the most common names included forms like Eleazar, Sabbataeus, Joseph, Dositheus, and Pappus—a noticeably different mix, with far more Greek names, than that of Palestine.⁴ So a writer working in the diaspora would not merely fail to copy the Palestinian list; he would tend to default to the names of his own world. The watermark is not just hard to forge—the forger’s own background actively pulls him away from it.

Third—and this is the subtlest point—it is one thing to include a few common names and quite another to reproduce their relative frequencies. An author could easily salt a story with “Simon” and “Mary.” What no ancient author had the tools to do was ensure that the second-commonest name appeared about as often as the second-commonest name actually did, and that the rare names stayed rare in roughly the right proportion. There were no frequency tables in antiquity. Reconstructing them took Ilan a modern scholar’s lifetime of cataloguing inscriptions. The pattern can leak out of genuine memory; it cannot be engineered from the outside.

Premise 2: Do the Gospel Names Actually Match?

Here the numbers do the talking, and they are striking. Working from Ilan’s data, Bauckham found that the most popular Jewish male name in Palestine was Simon, followed by Joseph—and these are exactly the top two in the Gospels and Acts. In Ilan’s catalogue of Palestinian men, about 15.6% bore one of those two names; in the Gospels and Acts, about 18.2% of named men do.⁵ Widen the lens to the nine most popular male names—Simon, Joseph, Lazarus (Eleazar), Judas (Judah), John (Yohanan), Jesus (Joshua), Ananias (Hananiah), Jonathan, and Matthew (Mattathias)—and the fit holds: roughly 41.5% of Palestinian men in the data carry one of these nine, against about 40.3% in the New Testament narratives.⁶

The women’s names tell the same story. The two most popular female names in the Palestinian record were Mary (Mariam) and Salome; in the Gospels and Acts these are again the two commonest. The fit is looser here—about 28.6% of women in the external data bear one of the two, against about 38.9% in the New Testament—but the rank order and the broad shape match.⁷ The sheer profusion of Marys in the Gospels (the mother of Jesus, Magdalene, the sister of Martha, the mother of James, the wife of Clopas) is not a narrative embarrassment to be explained away; it is precisely what a real first-century Palestinian sample looks like.

The most elegant confirmation, though, is not a percentage but a behavior. Because a handful of names were so over-common, real first-century Palestinian Jews had a practical problem: in any crowd there were too many Simons and Josephs to keep straight. Their solution was to tag people—by a patronymic (Simon bar-Jonah), a hometown (Joseph of Arimathea, Mary Magdalene), a nickname (Simon the Zealot, Simon the Leper), or a trade. The decisive observation is that the Gospels apply this disambiguation selectively, and apply it to exactly the names that needed it—the over-common ones—while rare names are usually left bare. That is not the behavior of a writer sprinkling in local color. It is the unconscious behavior of a culture that genuinely had too many Simons.⁸

Two independent controls strengthen the case. First, the same profile shows up in Josephus, a first-century Palestinian Jew whose works name hundreds of his contemporaries—an entirely separate sample that yields the same top names and ratios, so the Gospels are matching not one flawed list but a convergent picture.⁹ Second, and crucially, the Gospels match Palestine specifically. They do not match the diaspora profile. If these were late Greek compositions written far from Judea, that is the last thing we should expect.

The Strongest Objections, Stated at Full Strength

This argument has drawn serious, technically sophisticated criticism, and it deserves to be met at its best.

The deepest recent challenge comes from Kamil Gregor and Brian Blais, in a 2023 article in the peer-reviewed Journal for the Study of the Historical Jesus, “Is Name Popularity a Good Test of Historicity? A Statistical Evaluation of Richard Bauckham’s Onomastic Argument.”¹⁰ Their case is not the lazy “so what, they used Jewish names” dismissal; it is a genuine statistical critique. They argue, first, that the New Testament sample is simply too small for the law of large numbers to do reliable work—a few dozen named people cannot establish a tight statistical fit. Second, they identify what they regard as anomalies: certain names appear in the Gospels at frequencies that, on a closer look, sit awkwardly with the external data. Third, and most pointedly, they contend that Bauckham’s thesis of eyewitness sourcing offers no statistical advantage over a rival model they call “anonymous community transmission”—a story in which the names accrued through a Palestinian Christian community over time, with no need for named eyewitnesses behind them. If a non-eyewitness model predicts the same data, the name fit cannot be evidence specifically for eyewitnesses.

A second, older objection—and the one that bears most directly on this essay’s narrow claim about provenance—comes from the Oxford scholar Christopher Tuckett, in his review of Bauckham’s book. Tuckett does not deny that the names match Palestine; he calls the book a “tour de force.” His point is sharper. Bauckham wants to conclude that the names could not have been invented from outside Palestine. But, Tuckett notes, what that conclusion really requires is evidence that people outside Palestine were unaware of Palestinian names—and that is exactly the evidence we lack.¹¹ Jews were mobile; Palestinian Jews lived in Rome and Alexandria, pilgrims streamed to Jerusalem, and Josephus himself wrote his Palestinian narratives in Rome for a Greco-Roman audience. The naming knowledge plainly traveled. So a sufficiently well-informed insider—a diaspora Jew steeped in Palestinian tradition—might in principle have reproduced the right names without being a Galilean eyewitness.

A Reply—and What It Can and Cannot Carry

Take the statistical objection first, because the field has now litigated it directly. In 2024 the same journal published a detailed reply by Luuk Van de Weghe and Jason Wilson—the latter a working statistician—arguing that Gregor and Blais’s negative verdict followed from questionable testing choices and data selection.¹² Using a chi-squared goodness-of-fit test, they found that the name occurrences in the Gospels and Acts fit their Palestinian context at least as well as the names in Josephus do, and markedly better than names drawn from ancient fiction or modern historical novels. It should be said plainly that Gregor and Blais have in turn replied, and the statistical exchange is ongoing; the weight of this essay therefore rests on the narrow provenance claim, not on the verdict that “the statisticians settled it.” That last comparison is the telling one: the appropriate question is not “is the fit perfect?” (no real sample is) but “does it behave like a sample drawn from the population, or like a sample drawn from an author’s imagination?” On the published evidence, it behaves like the former.

Two honest concessions are nonetheless owed to the critics. The sample is modest, and Gregor and Blais are right that one should not oversell the precision of any single ratio drawn from a few dozen names; the women’s figures in particular wobble, and individual anomalies are real. And their methodological demand—that the eyewitness hypothesis be tested against rival explanations rather than simply asserted—is exactly the right demand. This remains a genuinely contested area of scholarship, not a settled one; the disputed hinge is precisely how much the statistical fit can bear. What the reply establishes is more modest than “Bauckham wins”: it is that the fit is real and is not plausibly the product of invention. That alone is enough to carry the narrow conclusion of this essay, which is about provenance, not about proving named eyewitnesses stand behind each pericope.

And here the strongest objection should get the last word before the reply—and it is Tuckett’s, not Gregor and Blais’s. Tuckett would press: you have shown the names fit Palestine; you have not shown that only a Palestinian eyewitness could have produced them. Palestinian onomastic knowledge demonstrably circulated in the diaspora—Josephus is living proof—so a skilled, well-read insider writing in Rome could, in principle, have mimicked the pattern. To exclude that, you would need evidence of diaspora ignorance of Palestinian names, and by its very nature such negative evidence is almost impossible to supply. The onomastic data therefore underdetermines the eyewitness conclusion. This is a fair point, and it should be conceded cleanly.

The reply is to grant the logic and narrow the claim accordingly. Tuckett is right that the names cannot, by themselves, prove that a Galilean fisherman wrote them down. But the argument does not need that to do real work. Consider what the “skilled diaspora mimic” hypothesis actually requires: an author who not only knew which Palestinian names were common but reproduced their relative frequencies across dozens of incidental characters, and who deployed disambiguating nicknames on exactly the over-common names—all while having no frequency tables to work from and, so far as we know, no awareness that anyone would ever check. Josephus could write accurate Palestinian narratives in Rome because he was a Palestinian writing from memory; he is evidence that the knowledge travels with such a person, not that it could be reconstructed by a Gentile or a third-generation diaspora author from scratch. Tuckett shows the door is not bolted; he does not show that anyone of the relevant profile is known to have walked through it, or could have. The most economical explanation of an unconscious statistical fingerprint remains that it was left by people who actually lived in that world—whether the named eyewitnesses Bauckham argues for, or, at minimum, the early Palestinian tradition the names entered before the books were written far away.

The Honest State of Scholarship

Where does the debate actually stand? The underlying facts are not seriously disputed. No one denies that name frequency is local and time-bound (Premise 1), and few deny that the Gospel names sit far closer to the Palestinian profile than to the diaspora one. The Jewishness and Palestinian flavor of the names is broadly accepted across the spectrum.

The live disagreement is about strength and reach. On one side, Bauckham, Williams, and statisticians like Van de Weghe and Wilson hold that the fit is tight enough to point to early, locally grounded—and probably eyewitness—sources. On the other, Gregor and Blais argue the sample is too thin to license that inference and that alternative transmission models fit equally well; Tuckett grants the match but denies it isolates eyewitnesses specifically. This is a real, ongoing argument among serious scholars, now being conducted in the technical language of goodness-of-fit testing—a sign of the argument’s health, not its weakness. The responsible verdict is that the onomastic data is genuine and significant corroborating evidence for an early Palestinian provenance, while the maximal claim—that it proves named eyewitnesses—remains contested and should be held with a lighter grip.

Conclusion: What the Names Do and Do Not Establish

The argument from Palestinian names establishes something modest and durable: the personal names woven through the Gospels and Acts carry the unconscious watermark of early-first-century Jewish Palestine—the right names, in roughly the right proportions, disambiguated in the right way—and conspicuously not the watermark of the diaspora world in which skeptical theories often locate these books. That fingerprint is very hard to explain if the Gospels are late, distant, fictional compositions, and very easy to explain if their traditions trace back to people who actually lived among first-century Palestinian Jews.

But the limits must be stated as plainly as the strength. This argument does not prove that the named individuals are eyewitness authors—Tuckett’s caution stands, and the maximal version is contested. It does not, by itself, establish that any reported event happened, and it certainly does not establish the miraculous; accurate names are perfectly compatible with a story that later grew. It addresses provenance and dating, not inerrancy or the resurrection. What it does is relocate the whole conversation. It makes the popular picture—Gospels as late Greek legends untethered from Judea—very difficult to maintain, and so forces the real debate onto early, locally grounded testimony. For the fuller case that this testimony is eyewitness testimony, see the companion essay on the Gospels and eyewitnesses; for the wider pattern of incidental Palestinian accuracy, see the Jewishness of the Gospels and the case from undesigned coincidences. The names are one thread. But they are a thread the late-legend hypothesis has a remarkably hard time cutting.