You might have heard this one before, but it bears a revisit. Once long ago William Lane Craig started using the argument that a mainstream historian in the early 1960s named A.N. Sherwin-White had demonstrated (in Roman Society and Roman Law in the New Testament, pp. 189-93) that (in Sherwin-White’s words) “even two generations,” or roughly forty to eighty years, “are too short a span to allow the mythical tendency” of a story’s embellishment “to prevail over the hard historical core of the oral tradition” that might then be found preserved in a text, like the Gospels, therefore (Craig goes on to argue) we can trust that the Gospel narratives of the resurrection of Jesus contain lots of historical facts (which means, conveniently, all the details Craig wants to be facts, depending on whatever argument he’s making at the time). Many an apologist has aped his argument since. And Craig still uses it himself (you can find it on pp. 190-91 of his still-current apologetics manual On Guard, for example).

Yet almost none of what Craig said is correct or true. It’s already weird that these guys can only find one unpeer-reviewed lecture from a long-dead historian from fifty years ago to back their argument. That conjunction of details should send up a red flag right away. Always check the dates on how old and moldy and hard to find their outside “sources” tend to be; and what they actually say, and in what actual context, and with what actual merit—and what historians in the intervening lifetime have concluded about it since. A single vague lecture from 1960 does not establish a very strong foundation for the extraordinary claim the likes of Craig are making here. But Craig and his imitators all get wrong what Sherwin-White meant, reason invalidly from it, and incorrectly claim Sherwin-White “proved” his statement about rates of legendary eclipse (when in fact he didn’t; and, it turns out, plenty of evidence refutes it).

The Unkillable Bad Argument

I thought I’d killed this argument all the way back in 1999, and in further revision in 2005 when I walked it down across three whole pages in The Empty Tomb (pp. 168-70, and further refuted across pp. 170-82). And then I was certain Kris Komarnitsky had totally done it in in 2013, with his own take-down in Myth Growth Rates and the Gospels: A Close Look at A.N. Sherwin-White’s Two-Generation Rule for the online trade journal The Bible and Interpretation (I recommend reading Komarnitsky’s excellent treatment). Yet that same year, in Christian Origins and Greco-Roman Culture, Tony Costa would still be making the same false argument, this time in defense of nearly every other miracle attributed to Jesus (pp. 142-43):

The analysis of Roman historian A.N. Sherwin-White … contends that even two generations is too short a time to erase the historical core of a recorded event, and he places the New Testament Gospels in this category. If [he’s] correct, this further supports the view that the healings and exorcisms of Jesus were part of the authentic Jesus tradition.

That’s a non sequitur, of course. Sherwin-White never said his notion of a surviving core of facts included miracles and wonders—and accordingly he never endorsed any as factual. It’s unlikely he’d agree with Costa. And no mainstream historian does; Costa’s conclusion is pretty much only that of a devoted Christian believer. To the contrary, Sherwin-White was only talking about mundane, procedural details of the trial of Jesus; and his conclusion was not that we could trust them all, but that we could trust some. Further work was needed to establish any as historical.

And that’s the first problem with all Christian apologetical abuse of the writings of the long-deceased Sherwin-White. The needed “further work” is actually unlikely to support either Craig or Costa. The claims they want to treat as factual are uncorroborated elsewhere (repetitions or embellishments of them are later and derivative, not independent texts). They are implausible. They fall into a category of ancient stories that were often untrue. And documents from decades earlier by a contemporary of the events in question that lack such narrative mythologization—the letters of Paul—contradict Craig and Costa. Paul’s account of the nature of the resurrection and how it was witnessed or confirmed does not agree with the Gospels, and Paul outright said Jesus abandoned all supernatural power in the incarnation—and accordingly never once mentions Jesus ever having been a miracle worker at all, much less a healer or exorcist. That bodes ill for those details being authentic in decades-later mythologizations. But the point is, one cannot argue from Sherwin-White’s point about myth not erasing all core details, to accepting just any details you want, much less the most incredible of them.

Even more absurdly, in 2014, Christian apologist Gregory Monette, in The Wrong Jesus, claimed “A.N. Sherwin-White concludes that it takes longer than two generations for legendary embellishments to wipe out the core of historical facts” and “this protects the entire New Testament” (p. 178). The entire New Testament! Talk about going completely off the rails with illogical, hyperbolic inferences. There’s a reason I think these conservative Christians are delusional. Likewise, in some 2013 podcasts, Christian apologist Bill Pratt claims saying “the resurrection narratives are mythical flies directly in the face of Sherwin-White’s analysis,” on the head-scratching logic that “the resurrection” means the same thing as “the Gospel narratives of the resurrection.”

I doubt Sherwin-White would ever have endorsed such an equivocation fallacy. Certainly no honest historian would confuse a general abstract belief with every single detail in a later legend produced by that belief. That “Christians believed in the resurrection” was a core fact is indisputable; but that’s not at all the same thing as the later Gospel narratives being built out of core facts. They are, rather, myths invented to convey that single core fact of what Christians believed about Jesus and what it meant. There is no evidence any of those details came from any actual source, or were ever literally true. This confusion plagues the thinking of even prestigious Christian apologists. There’s no evidence it plagued Sherwin-White. He is more likely rolling in his grave.

Yet it persists. I searched online for all appearances of this “Argument from Sherwin-White” published in just the last year or so, and I received pages and pages of relevant hits. A 2019 article by Dean Meadows asserts Sherwin-White said “that it takes far more than one lifetime” for “a legend [to] develop and spread” (that’s not what Sherwin-White said, nor anything that logically follows from anything he said). A 2019 article by Kevin Steele asserts Sherwin-White’s claim entails “any evidence to the contrary would have been preserved in the annals of history” (that’s also not what Sherwin-White said, nor anything that logically follows from anything he said). A 2019 sermon by Levi Lusko claims Sherwin-White “studied the rate of development for a legend” (he didn’t—this oft-repeated notion that Sherwin-White’s remark was based on “a study” is itself a legend) and, we’re told, “he concluded” from that non-existent study that “not even two generations is enough time for this kind of mythology to develop” (which is also not what Sherwin-White said). And, again, this 2019 interview of Lee Strobel by Tony Nochim shows Strobel confusing the Corinthian Creed with the Gospel narratives (a common apologetic trick), and then misusing Sherwin-White’s claim to bolster the latter—yet what Sherwin-White said would, even if true, only support the bare unnarrated facts of that creed but wouldn’t support the details of the much-later resurrection stories.

And that’s just a sample.

Back to Reality

In actual fact, Sherwin-White never conducted any study of this question, he never said myths and legends couldn’t arise in two generations (or even immediately), and what he did actually say—that some kernels of history must survive even in a highly embellished legend two generations later—is not only demonstrably false, he based his assertion of it on a single example that didn’t even evidence the point. This has been pointed out a lot. Not only, as I just noted, by myself and Komarnitsky, but many others—including multiple times by Vince Hart (in 2007, 2010, and 2011), and recently by Bob Seidensticker.

As I wrote before, if you read Craig’s original version of the Argument from Sherwin-White, to someone unfamiliar with what Sherwin-White actually wrote it would certainly seem as if he believed that no part of the Gospel stories is legendary, that for legends to appear in them is “unbelievable” (despite Craig’s use of quote marks, a word Sherwin-White never used), that “tests” (plural) have been performed on the text of Herodotus (a single historian—a detail since dropped in most iterations of the myth now), and that these tests “show,” convincingly enough to use the argument at all, that legends require many generations to develop. But in the actual text Craig cites, Sherwin-White never argues any of these points. People have caught Craig lying a lot. This is just another example. Yet this one spawned a still-living legend roaming the land of Christian apologetics like a spiritual brain-eating zombie.

The only saving qualification I can offer is that Craig did say those “tests” that never happened prove only that it is impossible for legends to wipe out some undefined hard core of historical facts, not (as many later versions of the myth assert) for elaborate legends to arise. But even this entails a misrepresentation of Sherwin-White. For Craig uses his argument to defend particular details of very marvelous narratives that Sherwin-White might not have considered the “hard core” of facts. For instance, Paul never mentions anyone ever finding an empty tomb; that detail appears more likely to have been invented by Mark; and Paul appears to make clear the apostles only saw the risen Jesus in fleeting dreams or visions or ecstasies, not in any of the ways the Gospels contrive, and certainly not having dinner with a reanimated corpse for weeks on end. The hard core of the story may simply be the bare fact that “the apostles said they saw the risen Jesus,” not how they said they saw him, or even how or even whether they actually did.

All that Sherwin-White actually objected to was the notion that “the historical content is…hopelessly lost.” Not to the notion that it was buried under myth, legend, and fabrication—or as Sherwin-White himself concedes, “myth,” “embroidery,” “propaganda” and “distortions,” since he was never talking about any miracle narratives but only mundane political and legal facts and thus wasn’t even talking about the ways marvelous tales arise. Sherwin-White totally agreed the Gospels were littered with mythmaking and fabrication. He wasn’t arguing it wasn’t, much less that it couldn’t be. He only meant there must be some facts hidden in the mix that we can extract with the right tools, a view long popular in Jesus studies that has actually since come under severe doubt. Indeed, every formal study conducted of the “tools” proposed to extract facts from the Gospel legends has concluded those tools don’t work. Consequently, historians are starting to doubt a great deal more in the Gospels than Sherwin-White’s generation did (see my summary of the scholarship in Chapter 5 of Proving History). So citing Sherwin-White even on what he did say is simply to lean on long-obsolete opinions since widely overturned. Admittedly, a typical behavior for Christian apologists.

But worse than the way apologists have distorted what Sherwin-White said, is the fact that even what Sherwin-White said was not supported by any evidence. He did not conduct any “tests” or complete any “study” of the matter. He simply offered a single example of a single sequence of events in a single historian, Herodotus. A sample size of one cannot establish a generalization, so Sherwin-White’s method was fatally defective and incapable of sustaining the assertion he based on it. Even worse than that, his sole example didn’t even evince what he claimed.

Of course, when you hear it from Christian apologists, you never hear that Sherwin-White admitted that even in Herodotus there has been “remodeling” and “a receptiveness to falsehood” and often a lack of “detached criticism,” and the use of allegory and symbolism, as well as a motivating bias to shape the stories toward an agenda. Sherwin-White merely claimed we should assume the result has not been mythologized out of all recognition. This is a substantially weaker conclusion than Christians using his lecture represent it. You will find plenty of legendary material being believed, repeated, or created by Herodotus or his audience, thus demonstrating that the same thing certainly could have happened in the Gospels—exactly the opposite conclusion Craig and his imitators are claiming we should draw from what Sherwin-White said. Which illustrates the dishonesty, irrationality, or basic historical incompetence that characterizes Christian apologists, which ought to undermine anyone’s confidence in their conclusions about anything.

Indeed Sherwin-White’s methodology was hosed from top to bottom. For example, he even fails to note the crucial differences between the Gospels and Herodotus—in quality, motives, content, identified sources, critical stance. The Gospels were religious propaganda composed by unknown authors in a foreign language and a foreign land, naming no sources for anything; Herodotus was a named critical historian, with no comparable religious agenda, writing about events in his own land and language, who often states his sources and evaluates the merits of what they told him. So no parallel can honestly be drawn between these texts the way Sherwin-White would wish. Fact is, we have no critical histories about Jesus or even Christians, written by any contemporaries, or even two generations out. Thus we cannot say what such histories would have contained. You cannot argue from either the contents or the silence of documents you don’t have. The Gospels are a breed apart as sources. Far less reliable than Herodotus. And Herodotus is universally recognized to be fairly unreliable.

Ignoring these crucial differences, and the impossibility of reaching general conclusions from single examples, Sherwin-White contends that “Herodotus enables us to test the tempo of myth-making, and the tests suggest that even two generations are too short a span to allow the mythical tendency to prevail over the hard historic core of the oral tradition,” the obvious origin of Craig’s characterization of his argument. Including the hyperbole of the plural. Sherwin-White’s “tests” consist of nothing more than one single example. No plural. No “tests.” And that one example doesn’t even evince what Sherwin-White falsely generalizes to.

That single example (in Histories 6.120-123) is a legend we already know was circulating in the time of Herodotus, yet Herodotus claims he shall recount the “truth” rather than the legend. And we know Herodotus had not just oral lore but a contemporary official inscription to check the legend by, a condition never available to the authors of the Gospels. Of course this is not evidence against the rapid creation of a legend—for the legend had already arisen, and indeed in less than two generations. As Sherwin-White concedes. But it’s also not evidence legends can’t rapidly eclipse facts—it is merely evidence that it didn’t happen on this one occasion; and only because Herodotus was, unlike the authors of the Gospels, a persistent skeptic, and had sources that would never have been available to the authors of the Gospels or their readers or even critics. In fact, Sherwin-White’s example actually teaches us we should distrust the Gospels, for they represent the “version of events” maintained by those devoted to the myth—and we have no Herodotus to correct that record by.

Thus, what Sherwin-White presented only a single item of evidence of, was the relatively critical mind of Herodotus, a stance entirely lacking in the Gospels. It could even be evidence of his lack of any relevant agenda or bias in this particular case, a lack of motivation once again entirely lacking in the Gospels. So this tells us nothing about the Gospels—the analogy is not even portable. This “example” is also a story about something that happened in the very city in which Herodotus is writing. No Gospel can claim even to have been written in the same country as the events it relates. Thus, as I show in the conclusion of my old article (updated in my chapter in The Empty Tomb), contrary to Sherwin-White’s single, irrelevant, cherry-picked example, the numerous patent absurdities and falsehoods Herodotus relates with minimal skepticism proves the Gospels should be expected to be much much worse. But you won’t hear Christian apologists admit this. Instead you’ll just hear William Lane Craig’s endlessly repeated lie, that Sherwin-White proved “there simply was insufficient time for significant accrual of legend by the time of the gospels’ composition” therefore “skepticism with regard to the appearance traditions in the gospels” is “unwarranted.” Which is not anything Sherwin-White said, or would have endorsed, or even been able to support with any evidence, much less the only paltry evidence he actually presented.

How History Actually Works

There’s a reason to be suspicious that Craig and every apologist since keeps leaning on this one, obscure, old-timey lecture: historians often make unevidenced or implausible assertions that, when checked, later historians expose as such, or simply reject. History as a field is continually self-correcting toward greater truth and accuracy over time. So not only should one have actually examined and thus noticed Sherwin-White’s argument is pretty lousy, but one should also have asked what later historians thought about Sherwin-White’s dismal argument.

Komarnitsky dug up one such example, quoting Peter Brunt from 1964:

Sherwin-White has done me the honour to cite a comparison I drew with our accounts of Alexander whom some of his own contemporaries treated as a god….[It is true that Alexander’s history was still able to be written,] but Alexander’s career was public in a sense which that of Jesus in Galilee was not. … If the synoptic Gospels reflect traditions that grew and were remoulded in the changing experience of the Palestinian Church, how can we objectively distinguish between what is original and what is accretion, seeing that the Gospels themselves must be almost our only evidence for that changing experience? … Sherwin-White has not provided, as he thinks, conclusive reasons to reject the view … that the history of [Jesus’] mission cannot be written.

And as Brunt went on to point out, exactly as I just did, Herodotus and the Evangelists are not at all comparable in the way Sherwin-White’s argument requires. They share neither methods, nor aims, nor skills, nor means. And as Komarnitsky himself points out:

[T]he Gospels are an understandable exception to what classical historians normally deal with, because classical historians rarely if ever deal with the written records of a highly revered religious figure who had very little contemporary significance to anyone but his followers when he was alive and to his worshipers after his death and where the entire written record comes only from those who worshiped him. Because of this, using the myth growth rates observed in other ancient records as a baseline to say what should be observed in the Gospels is a mistaken approach.

I’d go one further. When we do get those rare comparands (fawning hagiographies of otherwise little attested figures, etc.), we find exactly the opposite results from Sherwin-White’s: such texts are the most absurdly prone to exaggeration and invention, to the point of leaving nothing discernibly factual in them—even if anything in such stories happened to be factual, we have no way to know which details those may be. Without more reliable external corroboration. Which we simply don’t have for any of the things Craig and his imitators want. And that’s even for texts produced within a single generation.

Thus Brunt later went on to agree, concluding that Sherwin-White’s remarks “do not convince me that he had deeply considered this whole matter.” He was instead simply “a practising Church-man, and this may explain his unconvincing adventure into apologetics.” Numerous other historians (in fact a majority of those who reviewed Sherwin-White’s book) concurred with Brunt. As Komarnitsky quotes J.J. Nicholls, “the discussion, as far as it goes, is interesting, but it is too sketchy to be convincing.” And that was that. The claim has never been repeated by any historian since. Which should be obvious. Had any historian repeated the assertion with any confidence, Craig would be quoting them. Instead, he had to dig half a century back to get a long-abandoned, badly argued assertion to contrive his own argument on. Welcome to Christian apologetics.

Discover more from Richard Carrier Blogs

Subscribe now to keep reading and get access to the full archive.

Continue reading