David Marshall’s Bizarre & Dishonest Defense of the Historicity of Jesus

Liars disgust me. And David Marshall flat out lied about my work on public radio. He should be ashamed. But Christian apologists rarely are. They lie with impunity. The ten commandments be damned. Today I’ll briefly discuss that show, then detail what’s wrong with Marshall’s awful book defending the historicity of Jesus.


I really like Justin Brierley as a host. I’ve done his Unbelievable show several times now, which airs on Christian radio in London and beyond. He’s sharp, asks good questions, quickly gets the key points, and focuses the discussion. This time I was invited to engage David Marshall on his new book Jesus Is No Myth! The book is absolutely face-palmingly awful. But I’ll get to that later. The focus of this show was to be on whether the Gospel Jesus is based on other literary and religious characters, although we didn’t dive as much into that as we could have. What ended up being the main themes are the criteria Marshall thinks render the Gospels historically true accounts and not literary constructs; plus a little on why I think we can doubt the historicity of Jesus. But much of it became about how Christianity was a product of its time and culture, and not particularly unique (any more than every religion is unique).

Marshall’s entire schtick, on the show and in his book, is that in literary analysis similarities can all be dismissed whenever there are differences. So, by his logic, Westside Story cannot possibly be based on Romeo & Juliet because Westside is a vulgar-tongued musical about modern city gangs, and Romeo is an old high-dialect prose drama about medieval royalty. And Jesus can’t be based on Elijah because 1 & 2 Kings is a chronicle and the Gospels are not. And Jesus can’t be based on Moses because Jesus didn’t carry a staff. And pretty much any silly nonsense like that you can imagine. These aren’t things Marshall said; they are just what follows from his methodology. Needless to say, Marshall’s method is not accepted or used by any professional in history or literary studies. And indeed he cites no peer reviewed source as using or recommending the method. He just made it up. (More on that later.)

I won’t summarize the whole exchange. You can listen yourself at Premier. Many of the points I made I’ll cover in my book review below. But one theme that came up is how dishonest Marshall is in his presentation of data. Like James McGrath, Marshall lies. Marshall even covertly leaned on McGrath’s lies in the show, by magically multiplying that single Christian apologist into ‘many mainstream scholars’ who ‘panned’ my book. For the record, no academic review has ever panned my book; McGrath is the only person with a relevant Ph.D. who has ever even reviewed the book, and still never in any academic journal, and in every instance he lied repeatedly about the book’s content—and that’s not just a claim: I have extensively documented the fact. You should also note that I debated David Marshall before, and his epic loss was so embarrassing that he’s held a grudge against me ever since, starting even back then a campaign of lies and lunacy that is replicated in much of his new book.

For Marshall’s dishonesty, I gave the example of his book’s section attempting to argue that Christianity wasn’t a “Jewish mystery religion” (Jesus, pp. 20-24). There he lies by claiming I rested my conclusion that it was on just four generic criteria (false) and implying I was just making those criteria up (false). Then when I called him out on these falsehoods during the show, he not only stuck to these lies, he added an even more shameless lie to top them off: that I never used more appropriate criteria like the role of secret doctrines (fantastically false). Anyone who is going to lie like that, effectively to my face, in public, before an audience of thousands, and not even correct or apologize for it, is disgusting.

To illustrate what I mean, Marshall asks in his book “Why does Carrier focus on these four points?” (p. 21), as if there were no answer—despite his knowing full well the reason is that these criteria were developed and published by peer reviewed experts on the mystery religions. I did not make them up; I draw them from Petra Pakkanen’s peer reviewed book Interpreting Early Hellenistic Religion, which I plainly cite and explicitly say I’m drawing these criteria and my analysis from (OHJ, pp. 103-06). Marshall doesn’t tell his readers these criteria come from the expert peer reviewed literature; instead he implies I just made them up, and not only that, but that I did so for a nefarious reason, that I was deliberately “ignoring [the] essential traits” of mystery religions “and focusing on accidental ones” to deliberately trick people. Again, he doesn’t mention that he is thus accusing Petra Pakkanen and the Finnish Institute at Athens of tricking people with frivolous criteria. Marshall. Who, unlike Petra Pakkanen, has never published a single paper under peer review on the mystery religions and has no relevant graduate training in Greco-Roman religion.

That lie is bad enough. And it exemplifies his whole book, wherein throughout he almost never interacts with any peer reviewed literature (other than mine), neither to defend his own invented criteria, nor when critiquing anyone else’s arguments. Like here, he completely ignores Pakkanen and her academic monograph; he even deceives his readers into believing she and it don’t exist. But it’s worse that he then harps extensively on how it’s absurd of me to conclude Christianity is a mystery religion on just these four generic criteria. That is not just a lie. It’s a damned lie. In OHJ my section establishing Christianity was a Jewish mystery religion, labeled “Element 11,” begins on page 96. I survey ten paragraphs full of other criteria spanning seven entire pages establishing that conclusion before I even get to the four Pakkanen critera! Those are just the last four of a long line of other criteria (I discuss the Pakkanen criteria on pp. 103-07, after having surveyed more essential criteria on pp. 96-102). So by telling his readers I only used those four, David Marshall is flat out lying.

His dishonesty doesn’t even end there. Because he not only harps on my only using those four criteria (a damned lie), he makes an issue of how generic they are, that they aren’t the “essential” criteria for a mystery religion. Note that in his book, he still never tells his readers what those “essential” criteria are. In fact, I extensively summarize and apply every such essential criterion he could possibly mean (including the role of teaching mysteries, and a ritual baptism to secure eternal life, by the agency of a suffering savior, who is always the son or daughter of God, and the sharing of sacred meals to commune with their Lord and Savior: OHJ, pp. 96-102). So on the show I asked him what he thinks I left out. He said I never discuss in my book the most essential criterion of all, the role of secret knowledge (the core notion of the “mysteries”). That is a damned fucking lie. A Christian bearing false witness, pissing right on his own ten commandments, and right in the face of Moses.

Not only do I discuss the role of secrets and “mystery” concepts in early Christianity in my section establishing its status as a mystery religion in OHJ (pp. 96-98), I even devote an entire additional section to the fact that early Christianity employed secret doctrines just like other mystery cults did (“Element 13,” pp. 108-114). So the extent of Marshall’s dishonesty here is so vast it would shock even Donald Trump. It’s well enough to thoroughly discredit him.

Marshall’s shameless willingness to lie on public radio is appalling. But that he lies already so extensively about this in his book is reason enough not to trust anything else in it. If he is being this dishonest here, where else is he also lying about the things he talks about? The book is rendered useless by this fact. You may as well light it on fire and cook hamburgers over it. (Notably, Marshall has also harassed and lied about Matthew Ferguson, whom he also dishonestly criticizes in this book: see More Lies, Polemics, and Vehement Language from Christian Apologist David Marshall.)

This also illustrates a general reality. Once again, critics of On the Historicity of Jesus can only denounce it by lying about what it argues. Again and again, that’s the only way it ever gets treated. This all but establishes historicity is indefensible. Because if it could be defended honestly, it wouldn’t have to be defended with lies. When that’s all anyone can think to do to argue against it, even after two years of opportunity to find any actual relevant error in it, it’s time to stand up and take notice.

The Godawful Book

Marshall’s Jesus Is No Myth! is full of crappy and illogical arguments, even bizarre crankery. It has no coherent organization, rarely cites or addresses peer reviewed literature in any substantive or accurate way, and never once describes my book’s theory or the argument I make for it—nor does he do so for any theory of ahistoricity, rendering his book completely useless for its one stated purpose: to defend historicity against its critics. Thus by any standard, garbage.

  • Lies

His book contains a constant stream of lies, of course. Not just the one I just documented. He also lies about me using Apollonius of Tyana (and Philostratus’s “Life” of him) as a parallel to Jesus (pp. 199-211). I never once do so anywhere in OHJ. He quotes me supposedly doing so in our debate, only by taking my remark out of context; but he is supposed to be addressing my peer reviewed case against historicity anyway, not casual offhand remarks. Anything else is a blatant straw man fallacy.

In this month’s radio show Marshall balked at my suggesting that the Gospels have anything in common with Aesop’s Fables, even though I didn’t say the Fables, I said the Lives. Despite claiming to have read my book, he is evidently mysteriously unaware that it contains a whole numbered section on the parallels between the Gospels’ narratives and themes and those in the Lives of Aesop (again, not the Fables: “Element 46,” pp. 222-25). And indeed, in his book, where he claims to refute all parallels between the Gospels and other literature, he never once mentions my comparison with the Lives. So not only does he lie by misrepresenting the analyses he does try to denounce; he also lies by leaving out tons of analyses that undermine his book’s thesis, even though he is falsely claiming to have addressed what we’ve argued—instead, he conspicuously dodges plenty of what we argue, and doesn’t tell his readers what he’s evaded or left out.

His lies are also often compounded by idiocy. For example, he includes a lengthy and embarrassing rant against Frier’s Life Table, which exhibits average life expectancy in the ancient world using modern third world actuarial tables (which I then use in OHJ, “Element 22,” pp. 148-52), claiming I made up its contents, and that the table “can’t” be true because, he claims, the math in it entails crazy conclusions. In fact, Marshall is just an idiot who sucks at math. He makes egregious and embarrassing mistakes in reading the table, and his entire mathematical argument is one giant face-palm.

But the most shocking part of this is not his lousy math. It’s that he is, again, a fucking liar. He clearly intimates that I just made this table up, and that it’s mathematically impossible. In actual fact he well knows it comes from the peer reviewed demographics literature. He knows this because it says so, the sources fully cited, right atop the very page he is getting this from. That’s right. Frier’s Table is a real actuarial table built out of actual data taken from third world populations and published in standard peer reviewed demographics textbooks. If he thinks that table is impossible, he’s going to have to go argue with the actual professional demographers who built it. But more to the point, he conceals all this from his readers. That’s, again, lying.

I could continue listing lie after lie, but you get the point. Moving on…

  • Methods

Marshall’s method consists of finding “stylistic and literary qualities” in the Gospels that prove they are historically true and accurate accounts and not literary constructs. But none of the “stylistic and literary qualities” he enumerates are peculiar to historically true books; in fact they have no connection with how modern historians determine a source’s reliability. His method is wholly fabricated, and not even remotely based on how actual professional historians or literary critics do anything. Marshall also confuses method with purpose, and structure with style; especially when he keeps misquoting me as saying many other works share “all” the qualities of the Gospels—meaning not every single conceivable property (despite his lying about my meaning that), but only the properties I was discussing at the time.

All novels and myths have different messages and purposes and value schemes, and use different styles of presentation. So finding differences in those properties tells us little to nothing about the underlying truth of what they’re saying. What makes them the same in construct, meanwhile, is the method by which they are constructed. It doesn’t matter whether it’s in prose or poetry, high style or low style, first person or third person, or any other trivial aesthetic variation. They still use the same methods of communicating what they need: they are biographical narratives represented as true, yet they never name their sources, or never critically discuss them (contrast the example I exhibit from Suetonius in Not the Impossible Faith, Chapter 7); they are rife throughout with fabulous and improbable events; they follow a deliberate story arc from introduction to climax; they promote certain religious beliefs; and they do so through a central hero. None of the differences Marshall looks at change any of this.

Contrary to his extended red herring fallacy, what real scholars actually use to identify myth and fiction are those properties I just listed—especially the impossibly ubiquitous occurrence of fabulous events throughout the narrative. In Mark 1 alone, the sky tears open and a magical voice booms from heaven and angels bring Jesus food. In Matthew, the same story has grown: now Jesus flies through the air with Satan and observes the whole world from a mountaintop (the author of that story not being aware, apparently, that the earth is a sphere, or how geometry works). And that’s all before we’ve even gotten to the start of his ministry!

Elsewhere, Jesus magically withers a fig tree. He drowns two thousand pigs by casting spirits into them. He convinces several men to abandon their careers and families and follow him as disciples after showing up out of nowhere, doing nothing, and uttering just a couple of sentences. He clears a busy dozen acre temple bazaar, single-handedly, without the armed battalion on station lifting a finger against him. The sun goes out for three hours. An eighty foot temple curtain magically tears completely in two. Hordes of zombies descend on Jerusalem. Jesus has conversations with demons. Walks on water. Commands the weather. Magically conjures hundreds of fish and loaves of bread. Heals amputated ears.

Need I go on? This is myth. It’s exactly what myth looks like everywhere else. And Marshall himself would call it myth had he read these same exact stories from any other religion. He would laugh at anyone who tried to argue him into believing those stories, using the same bogus stylistic analysis he attempts to scam us with himself. He would tell them that their stories never cite their sources (nor critically evaluate them), look a lot like other tales, follow a convenient propagandistic story arc, and are full of fabulous and improbable events, and therefore, even Marshall would say, they are clearly not true or accurate historical accounts of anything. Likewise if he found all the same artificial structure and allegorical construction and emulation of prior myths (like those of Moses and Elijah) that I document in OHJ (e.g. pp. 396-506), none of which he ever even mentions in his book.

Though Marshall will never tell you, real historians identify myth not by any of the stylistic features Marshall enumerates, but by quite different criteria (OHJ, pp. 389-95). Which boil down to how much a narrative emulates prior myths and stories, how much it relies on fabulous and improbable events, and how little of its core features have any external corroboration (OHJ, p. 394). The more a narrative hits all three of those criteria, the more likely its content is to be mythical. The clearest test case of that general point, for which we have the most, and most consistent, data, is the Rank-Raglan test (OHJ, pp. “Element 48,” pp. 229-34), which Marshall attempts a ridiculous rant against (I’ll examine that below).

Of Marshall’s thirty enumerated criteria, almost none relate to determining the historicity of a narrative. They almost all relate to the authorial purpose in composing the story, which tells us nothing about whether the content they use to accomplish that purpose was historically true. The Gospels are counter-cultural fiction, and as such would be expected to do all the things Marshall marvels at (like criticize elite culture and values; make radical statements; claim prophecy has been fulfilled in unexpected ways) regardless of whether they did so with truths or fictions.

Truthfully, the Gospels contain teachings no more radical than anything else taught by ancient Rabbis, Sages, and Cynic philosophers (OHJ, “Element 32,” pp. 173-75); even the Beatitudes are just a reworking of prior Jewish traditions (such as we’ve found at Qumran: OHJ, “Element 33,” pp. 175-77). But even if they have anything original in them, that does not entail it came from Jesus. If Jesus could invent something impressively new, so could any Christian missionary. No honest historian uses the criterion that “no one can ever have innovated radical ideas but Jesus.” That just isn’t a marker of historicity—anywhere, for anyone. Other than Christian fundamentalists.

Of the few criteria Marshall relies on that bear any vague similarity to real criteria used to assess historicity, in not a single case does he ever rely on or address any of the peer reviewed literature regarding them, even though I cite and summarize that literature in Proving History (cf. Chapter 5), the book he claims to be responding to—and no one can advance a discussion in the field who simply flat out ignores the entire present state of that discussion. (As for methods, BTW, I could also mention his ridiculous crank Chinese etymology at the end of the book, which fully merits the classification of laughably bizarre, wherein he “proves” Jesus and the gospel were secretly hidden in ancient Chinese characters. But you don’t need me to explain why that’s delusional to the scale of hilarious.)

  • Criteria

Marshall attempts some dishonest and illogical defenses of the traditional criteria (Jesus, pp. 66-81, 89-97), but I needn’t examine that further. What I have already written in Proving History (Chapter 5, esp. pp. 124-69, “Criterion of Embarrassment,” and pp. 169-72, “Coherence,” and pp. 172-75, “Multiple Attestation”; also “Contextual Plausibility” and “Dissimilarity,” pp. 123-24, 137-38, 176) already refutes him. He also reveals considerable ignorance of classical literature—for instance, he seems not to know (in Jesus, pp. 68-70) that we have nearly a thousand epistles written by Cicero, by which we actually learn a great deal about his interests, beliefs, knowledge, sources, and concerns, precisely what we don’t have for any author of the Gospels.

And beyond outright ignorance, Marshall also deceives his readers by once again leaving out a great deal of the evidence and argument I present for my conclusions regarding these criteria, misrepresents what he does discuss, and does not address any of the peer reviewed literature I cite and rely on for those conclusions. (He doesn’t even tell his readers that numerous leading experts have come to the same conclusions I have.) And then he builds absurd and illogical arguments that I think any non-insane person can see through without any help from me.

Meanwhile, of his own criteria, most are irrelevant to historicity. He never adduces any data they are ever demonstrative of historicity in any other religious narratives. And they are not specially linked to truthtelling—all of them can be used in fiction. That Jesus acts as a realistic mouthpiece for Christian values (Marshall criteria (7), (9), (15), (17), (18), (19), (23)) is exactly what a fictional Jesus would do, because that’s exactly what the authors of the Gospels want to convey. And those authors were counter-cultural critics of elite values and attitudes, so that’s also exactly what we expect to find in their fictional portrayal of their hero (Marshall criteria (15), (20), (22), (23)).

Similarly, the Gospels served as handbooks for missionaries, supplying them with convenient models to use and refer to when evangelizing groups and individuals (e.g., Proving History, pp. 156 and 178: Marshall criteria (6), (9), (10), (11), (15), (21), (23)). And the authors wanted gullible readers to believe their fiction—until they were properly initiated and instructed in its real meaning (e.g., Mark 4:9-13; OHJ, “Element 13” and “14,” pp. 108-24: Marshall criteria (1), (3), (4), (5)). Some of those authors we can tell even used reference books to “pad” their accounts with “historical color” (and even then made mistakes when they did).

Fiction often has accurate geographical, historical, and cultural details. Sometimes it doesn’t. Sometimes it does and doesn’t. Like the Gospels: Marshall falsely claims the Gospels consistently get the geographic and cultural details right (Marshall criteria (3), (4), (5)), but that’s only sometimes true. Mark made several mistakes that Matthew had to correct (OHJ, p. 459); Luke used reference books to get many details right, but still made mistakes in his depiction of Roman census law and procedure, messed up the chronology of Jewish rebellions, and outright falsified numerous historical facts (OHJ, pp. 362-63). Indeed, a lot the Gospels depict is completely contrary to the known facts, from trial law to temple logistics and unrealistic speeches (OHJ, pp. 362-63, 381-82, 402-03, 425, 431-32, etc.).

Marshall is failing at a basic principle of Bayesian logic: if the item in question, even in conjunction with others, is just as expected on fiction as on history, its likelihood ratio favors neither. It’s therefore not indicative of anything. Sometimes, even, the data establishes exactly the opposite of what Marshall thinks. He offers the presence of incidental details in Gospel stories (though actually there aren’t very many) as evidence of historicity (Marshall criterion (8)). But studies of legendary development have confirmed that’s actually exactly what happens to legends: they accumulate more and more incidental details and realistic “color” over time (OHJ, pp. 480-81, n. 195). Contrary to how he thinks this criterion would work. It therefore might even be evidence of fiction, rather than history. But even at best, it doesn’t make historicity more likely.

Another example is Marshall’s claim that the Gospels are more likely historical because “people exit the story without making improbable reappearances just to tidy up the plot or give curtain calls to popular characters” (Marshall criterion (14)), but he gives no examples of this ever being a practice in any form of ancient fiction. He cites only Dickens as exemplifying this trend. Sorry, that’s almost two thousand years too late to be relevant. Trends in modern fiction have no relevance to trends in ancient fiction. To the contrary, in ancient fiction, what was typical is for characters to pop in out of nowhere, perform their necessary function for the story, and then immediately disappear, never to be heard of again (unless once again important to the storyline: e.g. John the Baptist, Judas, Nicodemus, Lazarus), even though that’s highly unrealistic. The three women in Mark 15-16 are an example (OHJ, pp. 421-22); likewise the mysterious vanishing of all the brothers of Jesus, Joseph of Arimathea, even Pontius Pilate! (OHJ, pp. 371-75) At worst, disappearing characters who only exist to move the plot is a device of fiction, not history. But even at best, they don’t make historicity more likely.

It only gets worse when Marshall seriously cites as historicity-evincing criteria the fact that Gospel miracles are all “realistic” (!) and that Jesus fulfills scripture (!) (Marshall criteria (29) and (30)). These are actually criteria of fiction in actual mainstream history as a professional field (Proving History, pp. 114-17, 177-78). A hero’s improbably ubiquitous fulfillment of prophecies and oracles is evidence the authors are fabricating the narrative, precisely to have that effect; it’s not accepted anywhere (outside irrational fundamentalism) as evidence that such an improbable conjunction actually happened (see Newman on Prophecy as Miracle). And there is no such thing as a “realistic miracle.” That’s an oxymoron. A phenomenon never documented to actually exist is by definition unrealistic. And the ubiquitous appearance of scientifically implausible events in a story is accepted by all non-crazy mainstream historians as evidence of fabrication. It is not ever accepted as evidence of historicity.

The funny thing, of course, is that this means Marshall thinks Jesus withering a fig tree for the illogical reason that it wasn’t bearing figs out of season is realistic. That Jesus drowning two thousand pigs by casting spirit entities into them is realistic. That Jesus viewing “all the kingdoms of the world” from a mountaintop is realistic (gosh, which mountain would that be?). That Jesus simultaneously riding both an adult and a baby donkey into Jerusalem is not only realistic, but evidence it’s true! (Good luck figuring out the logistics of that: OHJ, p. 459).

The rest of Marshall’s criteria are even outright false (the Gospels actually don’t merit them) as well as irrelevant (even a story that merits them is not thereby more likely historical). For example, he cites as a criterion of increasing historicity that the Gospels are very early (Marshall criterion (2)), but fiction can be written immediately. Time has no bearing on that. By contrast, very late texts can actually be more likely historical than early ones: Arrian’s history of Alexander (written 500 years later) is far more historically reliable than the inaccurate and legend-riddled account of Callisthenes who was an eyewitness! (On why Arrian is more reliable than the Gospels: OHJ, p. 22.)

Thus, nearness in time is not by itself a relevant criterion. But the claim is also false. The Gospels were written after even Paul was dead, much more any other known witness. The earliest can’t have been composed any earlier than forty years after the fact, by which time any witnesses there may have been would have been in their sixties or seventies, and statistically very likely dead by then (average life expectancy for adults was 48; and Christians had to endure persecutions, wars, and famines in those forty years, so their life expectancy was below average: OHJ, pp. 148-52). Accordingly, we have no evidence any witness was alive when Mark wrote (he refers to none; and no other sources attest to any). And that’s just Mark. Matthew wrote later still. And Luke, we know, had to have composed no earlier than sixty years after the fact, when any witnesses would likely be in their eighties or nineties, and almost certainly dead by then; and John wrote decades after that (OHJ, pp. 264-70). The Gospels were thus written suspiciously precisely when legends begin to run most rampant. And we have no evidence any witness was alive to gainsay them.

Similarly, Marshall claims the Gospels depict people and audiences reacting realistically (Marshall criterion (12)), but that’s both irrelevant (fiction can depict that, too) and false. The behavior of people in the Gospels is highly unrealistic. In the real world no one abandons their jobs and families and fanatically follows a total stranger the rest of their lives after hearing him utter two sentences. The occupants and managers and soldiers of the temple would not have sat idly by as Jesus thrashed the tables and merchants and animals there. The Sanhedrin would not have needed Judas to seize Jesus. Jews in Jerusalem would not have reacted to Stephen’s or Peter’s anachronistically ineffective speeches as they are depicted doing in Acts, because those speeches make no sense in that context, nor do the depicted reactions to them (OHJ, pp. 363, 381-84). Even the reactions to the miracles of Jesus are implausibly mundane. The disciples even implausibly forget he can miraculously invent vast quantities of food. There is actually hardly any scene in the Gospels that realistically depicts the actual complex and nuanced ways real people would actually behave had they seen or heard those things. This is actually evidence the Gospels are fabricated.

Likewise, Marshall claims the Gospels realistically develop the personalities of supporting characters (Marshall (13)), which is again both irrelevant (fiction can do that, too) and false. The motivations and character of Judas make absolutely no sense as depicted. Peter is a cypher. Why he vacillates as he does, what his aspirations are, pretty much anything we’d want to know about what sort of person he was and why he did what he did, is never revealed. And no one else is even developed enough to assess as a person. Everyone else in the Gospels is always in fact just a cardboard cutout who only ever speaks or acts so as to serve as a lesson for the reader or as a foil for Jesus or to move the plot, even when their behavior makes no sense or is wholly improbable for a real person. This is actually evidence the Gospels are fabricated.

The same thing goes for Marshall’s claim that the Gospels depict its actual historical characters realistically (Marshall criterion (16)). But that’s either not true or not known to be true. We know little about Caiaphas by which to assess his depiction as realistic; likewise for every other known historical person paraded into the narrative. Except Pontius Pilate. Who is not depicted realistically at all. Marshall tries to explain away the fact that Pilate violates rather than merits the criterion, by inventing reasons why he might do the things he does in the Gospels—but pointedly, it never occurs to the Gospel authors to give those reasons. That David Marshall knows how to write a more realistic depiction of Pilate than the authors of the Gospels did is not evidence the authors of the Gospels were writing history. It’s evidence they were writing fiction (OHJ, pp. 371-72, 374, 403).

Apart from his bogus “thirty criteria,” Marshall also relies on the crank “Argument from Undesigned Coincidences” (see Babinski and Ehrman), which is based on ignoring that Matthew and Luke are using Mark (and each other) as a source (just as Acts used the Epistles, and John used Mark and Luke), and rests on illogical premises about how authors and reality work. He also relies on the bogus claim that the frequency of names in the Gospels matches reality; n.b., it doesn’t: very common names (like Jesus, Lazarus, Ishmael and Manahem) are peculiarly less frequent than they should be; and the names that do appear are mentioned too infrequently to produce any statistically significant conclusion, or are disproportionately over-represented (like James and Phillip), or are actually conspicuously unusual for Palestine (like Nicodemus, Stephen, and Bartimaeus). The most typical names in fact all derive from the OT, which is evidence of symbolic emulation of the scriptures, not historical reality. Marshall cherry picks the evidence that conveniently confirms his name frequency hypothesis, while hiding the evidence that contradicts it. (And again, he sucks at math.)

Finally, it’s worth noting, of course, that several of Marshall’s “thirty criteria” are really just repeated examples of the affective fallacy: he is so moved by the teachings and personality of Jesus, he concludes it therefore must be true. Such reasoning would mean the Quran is true, and the Tao Te Ching as well, simply because Muslims (like, these guys) and Taoists (like, once upon a time, myself: Sense and Goodness without God, pp. 11-15) find it too moving and powerful to be anything but divine. The argument is thus shown to be invalid. Like almost all Christian apologists, Marshall sucks at objective reasoning. (And accordingly, his attempt to debunk Loftus’s Outsider Test for Faith, in How Jesus Passes the Outsider Test, is even more awful and dishonest than the book I’m presently reviewing.)

  • The Rank Raglan Class Again

Everyone is obsessed with the Rank Raglan Hero class. They are indignant that the Gospel Jesus could possibly have been based on a recognized and repeated hero trope. And they don’t grasp the mathematical reality establishing the existence of that trope. I’ve addressed most objections to its existence and application already (all ignored by Marshall, of course; again, his readers aren’t told what I actually argue) in On the Historicity of Jesus (Chapter 6; in conjunction with “Element 48,” pp. 229-34). I combined and updated the wording of the criteria to fit the actual matching data Rank and Raglan cited, and then I double-checked the scoring myself (and in result even downscored some members I believe Rank and Raglan overscored), and after all that, I found it held up.

Besides Jesus we have fourteen people in antiquity who match a score above 11 out of 22 on the Rank-Raglan scale, and not one of them is plausibly historical. That’s highly unexpected. There should be a lot of historical persons among those fourteen who got layered with that many legendary motifs, too, not just mythical people. That there aren’t, however, means that that just didn’t happen—or if it did, it happened only rarely. I conclude charitably that maybe as many as 1 in 3 people who got mythologized that much were nevertheless still historical. And I think that’s being way over-generous—indeed, that high a rate is not backed by any actual evidence (we have no evidence even a single member of the Rank-Raglan class was ever historical). So my conclusion is super-conservative.

This simply represents an obvious truth: that the more mythical features a historical character has, the more likely they are to be mythical. Anyone who would gainsay this has to produce evidence to the contrary: they have to collect all the highly mythologized persons in history, and count how many are historical, and how many not (or not plausibly). I am quite certain you will never get a result better than 1 in 3. The only reason I use the RR Class to do this is that it’s a very clear, consistent trope that is evidenced a remarkably large number of times (15 in all). But I could have tried a broader study, and collected everyone in history who is as mythologized as Jesus. That would include everyone who has all the mythical features piled onto Jesus: everyone who is all-at-once a worshiped deity, a mystery-cult savior, a dying-and-rising demigod, a culture hero, a heavenly founder, and so on (e.g. OHJ, “Elements” 31, 36, 46, 47; Chapter 6.1-2; etc.).

That class would have to include everyone who is:

  • A conveniently named godman (OHJ, pp. 239-42).
  • And cult founder (OHJ, pp. 8-11, 159-63, 557-63).
  • Who was not only a miracle working sage (OHJ, p. 230).
  • But also a preexistent incarnated being (OHJ, p. 230).
  • And a worshiped savior Lord (OHJ, pp. 96-108).
  • As well as a revelatory space alien (OHJ, pp. 137-41, 146, 197-206)
  • Appearing only in sacred literature (OHJ, p. 214-22).
  • Who dies and rises from the dead (OHJ, pp. 168-73, 225-29, 239).
  • And whose life improbably fulfills numerous prophecies (OHJ, pp. 141-43, 230).
  • And whose only biographies build him out of prior religious heroes he is meant to supersede (Moses, Elijah, Odysseus, Romulus, etc.) and are rife with fabulous and improbable events (OHJ, Chapter 10).
  • Which name no sources, discuss no sources, and have no known sources.
  • And are constructed from known counter-cultural hero narratives (OHJ, pp. 222-25, 430-31; PH, pp. 131-32) and popular ascending sage legends (OHJ, pp. 225-29).
  • And who becomes that mythologized in under forty years time (OHJ, pp. 248-52).

And for mathematical convenience I’ll not count here another telltale sign: those whose existence is wholly uncorroborated outside the myths of them (OHJ, Chapter 8).

If we then add the RR Class to that, the only person in history still in the class would probably be Jesus. Meaning Jesus is one of the most mythologized characters in the whole of human history. If we left out the RR criteria, we’d still find few members of the remaining class are actually historical. I very much doubt more than 1 in 3.

If you want to prove it’s a different frequency than that, you need to roll up your sleeves and do the work: collect everyone in history who is as mythologized as Jesus (as just enumerated), and count how many are actually historical. Until then, you cannot say the frequency is higher than 1 in 3.

It gets you nowhere to say “but a historical person could have had that much mythology piled onto them in just forty years.” That still doesn’t tell us how probable that is. You need a frequency: how often does that happen? Because that’s then the prior probability it happened to any such person, including Jesus. Marshall, like every other critic who attempts to avoid this conclusion, never says what the prior probability is, or how they know that, or what frequency data they base that on. They want to insist they “know” it’s a high frequency. But they never show any evidence that it is, or even that they are aware of any evidence that it is. So you have them, making shit up, backed by no frequency evidence whatever; and you have me, who ascertained a frequency from actual data. Who is winning this argument?

Of course, trying to get a different frequency from actual pertinent evidence is hard work, and Marshall is lazy; it’s also impossible, because guess what, even before you look at the list of persons as mythologized as Jesus, you can probably already predict the frequency of historical persons on that list isn’t going to be that high. So Marshall needs Jesus to not be in the Rank Raglan class. So he slews a slaw of ridiculous apologetic bullshit to try and get him out of there. But even outside the RR Class, Jesus sits at almost the most extreme degree of mythification that anyone claimed to be historical can even undergo; it’s thus not possible to argue against the conclusion that he should be among the least likely to be historical of almost anyone there is, totally regardless of whether he belongs to the RR Class.

I’ve discussed the facts and mathematical logic of the Rank-Raglan trope several times already, beyond what’s already covered in OHJ (see McGrath on the Rank-Raglan Mythotype; The Covington Review; The Tim Hendrix CritiqueThe Hallquist Review; and Two Lessons Bart Ehrman Needs to Learn). Here I’ll just quickly make fun of Marshall’s attempt to escape reality on this point (Jesus, pp. 28-37):

  • Marshall claims the Gospels never declare Jesus the heir of a king. Apparently David wasn’t a king.
  • Marshall claims the Gospels never declare Jesus the Son of God. [Roll laugh track.]
  • Marshall claims Jesus is never said to be reared by one or more foster parents. Because Joseph was actually the biological father of Je…oh no, wait.
  • Marshall claims we are told all about Jesus’s childhood. Because we are told about him as a baby and a young adult. Neither of which is his childhood.
  • Marshall claims Jesus never returned to his future kingdom. Apparently David didn’t rule Judea. And this didn’t happen.
  • Marshall claims the Gospels never declare Jesus King of the Jews. [Roll laugh track.]
  • Marshall claims Jesus’s ministry included wars and natural catastrophes. Because, Marshall insists with a straight face, drama is the same thing as wars and natural catastrophes.
  • Marshall claims Jesus did not proclaim any laws. Because, you know, his Sermon on the Mount was never enacted by a legislature.
  • Marshall claims Jesus was never forsaken by God. Even though Jesus said he was forsaken by God.
  • Marshall claims Jesus was never abandoned by the people. Apparently this didn’t happen. Or this.
  • Marshall claims Jesus was never driven “from the throne or city.” Because he doesn’t know what the word “or” means. Or that Jerusalem is a city.
  • Marshall claims Jesus’s rapid death was no surprise. Pontius Pilate was more experienced with crucifixion death rates.
  • Marshall claims nothing mysterious happened at Jesus’s death. Just the sun going out for three hours, an eighty-foot-high curtain magically tearing in two, a rock-shattering earthquake, and a horde of zombies descending on the city. Roman centurions were apparently more experienced with mysterious events.
  • Marshall claims Jesus’s children not succeeding him doesn’t count because Jesus “wasn’t a king” (right, Jesus totally wasn’t declared the King of the Jews…but his children didn’t even succeed him in running the church, either). And because he didn’t have any kids. Even though not having kids is one of the ways kids don’t succeed you.
  • Marshall claims Jesus has no tomb. Hmm.
  • Marshall claims defeating Satan in the desert doesn’t count as battling a great adversary. Even though Satan literally means “the adversary” and there was literally none greater.
  • Marshall claims that this contest didn’t happen before Jesus entered his future kingdom. Because he sucks at math, so he doesn’t know that Mark 4 comes before Mark 10.

If you aren’t laughing your ass off by now, you haven’t been paying attention.

As usual Marshall also lies, claiming for instance I only score Jesus on Matthew, not Mark (false: I note Mark scores Jesus at 14 out of 22 in OHJ, p. 232). And, again, he sucks at math (mistakenly claiming we need to know how many historical people don’t score above 11 in order to know how many who do are historical). But his attempts to downscore Jesus just make him look ridiculous. I hardly need any further argument.

The saddest thing here is that Marshall is totally fucking freaked out over this 1 in 3 prior chance of being historical derived from the Rank Raglan trend. It horrifies and terrifies him. He has to run screaming from it, or thrash at it with wild irrational pummeling. And yet that’s an extremely weak prior against historicity. It would be easily and quickly overwhelmed by any good evidence. Good evidence is something that’s, say, five times more likely to exist if the person existed than if they didn’t. A 5/1 Bayes’ Factor times a 1/3 Prior Odds gives you a 5/3 result in favor of historicity. Just from a single piece of relatively mild evidence (really good evidence has a Bayes’ Factor of hundreds or thousands or even millions to one; like the evidence we have for Julius Caesar). That defenders of historicity can’t even find enough evidence for Jesus to overwhelm a feeble 1/3 prior odds against it is what should be scaring them. The prior itself is so weak it’s not even scary at all.


You will never learn from David Marshall’s book what the peer reviewed case against historicity actually consists of, or what facts and arguments it actually rests on. You will never see any engagement with any of the independent peer reviewed literature that supports that case. You will never hear what the alternative explanations of all the evidence are (including the origination of Christianity as a sect and dogma), much less find any coherent explanation of why it’s wrong. You will have tons of evidence hidden from you. You will be lied to. And you will be distracted by irrelevant methodologies that have no basis or support in any professional field.

Indeed, Marshall’s entire approach rests on assuming miracle claims have the same probability of being true as mundane stories (e.g. Jesus, pp. 162-81). Of course his reasoning to that conclusion is totally illogical, and only betrays his unprofessional bias (see, by contrast, why real historians—even when believers like Raymond Brown—don’t act like that: Proving History, pp. 114-17). His book is thus just so much Christian apologetics, and not even good apologetics at that. It is not serious scholarship that would pass peer review, at least anywhere with respectable academic standards. It’s all smoke and mirrors, a continuous thread of handwaving with a bankrupt methodology, wanton dishonesty, and the concealing of evidence. I don’t recommend it. Other than perhaps for amusement. Or cooking hamburgers. Oh, wait, no. It’s probably toxic.



  1. Thanks for linking to my piece that questions the argument from the 1800s concerning undesigned coincidences, which is a pathetic argument founded on an early attempt to dismiss the evidence of literary connections and story development between the Gospels over time. Lydia McGrew by the way has composed a book advocating undesigned coincidences, published by a conservative Christian press. I may review it in future. Though she knows my opinion, she has not sent me a review copy yet. Perhaps she will.

    1. If you do review it, definitely let me know. I’ll signal boost that! You can also ask me questions when compiling it. I’d like there to be a good rebuttal on that. Or more than one even.

  2. While I certainly don’t agree with Marshall on the Rank-Raglan scoring, I’m not particularly convinced by your arguments either. Just to mention a few things:
    – Being a very distant descendant of someone does not make one heir of that person. Now I would agree that the Gospels imply that Jesus is, in some way, a heir of David, but then again that seems irrelevant since Rank-Raglan does not discuss ‘heirs’ in this sense.
    – Claiming Joseph as ‘foster parent’ of Jesus seems imaginative to me, since this requires us to read the Hebrew God as the *real* father of Jesus rather than merely the ‘reputed’ father. But this is not how Raglan treats this. For instance, Theseus is not reared by foster parents because Poseidon does not care for him, but because he is reared by his grandfather rather than king Aegeus.
    – As for Jesus’ childhood, well it depends on which sources we use. Of the canonical Gospels only Luke tells us anything, and not that much, but the Infancy Gospel of Thomas is early and extensive.
    – As for Jesus returning to his future kingdom, the problem is not that David did not rule Judaea, but that *Jesus* did not rule Judaea (or anything else). There are several cases of prophecy in the Gospels saying that Jesus *will* become king, but it never actually happens in the story (Christians obviously believe that Jesus will one day return to become a king, and we could construct an expanded Jesus story where we include this, but this would have massive effects on the RR scoring, for instance in this narrative Jesus is not losing favour or ever dying etc).
    – As for “wars and natural catastrophes”, again this is irrelevant to Rank-Raglan. Raglan describes the hero’s rule as uneventful, expanding that “..his story, from the time of his accession to the time of his fall, is as a rule a complete blank. The only memorial of his reign, apart from the events that begin and end it, is the traditional code of law that is often attributed to him.” I honestly have no idea where you ever got the ‘wars and natural catastrophes’ from.
    – The quote above also sheds some light on why it is quite problematic to equate a sermon with prescribing laws. Raglan very clearly referred to an actual body of laws such as used by judges. Raglan further writes: “As a fact, however, a code of laws is always the product of hundreds, if not thousands, of years of gradual evolution, and is never in any sense the work of one man.”
    – As for being driven from the city, Jesus is hardly driven from Jerusalem. He is simply executed, which technically is (just) outside the city walls, but that is not what any straight interpretation of ‘driven from’ would entail. And Jerusalem is not his city anyhow, so this requires massive shoehorning to make Jesus fit.
    – Having a ‘tomb’ is not the same as having a holy sepulchre which is what Raglan refers to. Now, Jesus has those too – today – but they do not feature in the Gospels. So either we have to allow much later sources, in which case we also have stories about his childhood, or we can’t score Jesus on this point.
    – The talk of an ‘adversary’ is again irrelevant, that is not what Raglan talks about but rather “the king and/or a giant, dragon or wild beast”. The event is a way for the hero to qualify for becoming king. Jesus on the other hand turns down a proposal for Earthly power, the purpose of that story is to show that Jesus is not at all interested in being a king, quite at odds with Raglan.
    – You place the supposed ‘victory’ before Jesus enters Jerusalem. But Jesus does not become king when he enters Jerusalem, rather he is executed. If we want to parallel the life of Jesus with the Raglan heroes, it seems to me that the clearest equivalent of ‘becoming king’ is the baptism by John. This has the form of a coronation event, and is how Jesus’ ministry starts. Since Jesus is never a king (in the story) in any straightforward sense, it makes most sense to replace ‘king’ with what he is: a preacher.

    1. You are treating two different scholar’s different lists as if it were one list and as if it were gospel. That’s invalid methodology.

      We can change their wording and criteria all we want, just as we do when we merge them into a single list. As long as more than ten people match a score more than half, we have a meaningful set. If that set is random, then most of its members should be historical, just as most historical characters are. So it won’t matter even if we just make up completely new criteria. The result will still reflect the actual prior for members of that set. Unless they are members of a set that is less frequently historical.

      So even if we take my wording of the RR criteria as a completely new set of criteria and not credit Rank or Raglan at all, the fact that it still has more than ten members and none are credibly historical rules out any chance the set I’ve created is a random sample of historical characters. Regardless of what Rank and Raglan were doing, my adapted criteria have also found a peculiarly ahistorical set of persons. I’m just being honest enough to give them credit for having first noticed this, even if they were a bit sloppy in their counting and formulation.

      That then becomes the prior that any member of that set is historical, just as it always would. The only way this frequency of historicity in that set would be purely chance accident is if there were far fewer members, allowing chance accident a respectable probability of producing the disparity. But with 15 members, that’s impossible, at the probability resolution I employ.

      So trying to treat Rank and Raglan’s wording like holy scripture makes no sense and serves no function. I’ve found a remarkable set. With so many members, it can’t be chance accident that all its members are agreed to be likely mythical. Therefore, members of this set necessarily have a high prior odds of being mythical. And that would be just as true if I invented all 22 criteria myself out of whole cloth. So trying to resurrect the dead scholarship and wording of Rank and Raglan is a waste of time. We need to attend to what I found, using my criteria, and my wording.

      That said, we have to score consistently…

      That Jesus was the Davidic heir to the throne is the entire point of the genealogies and the Gospels calling him the Son of David (three times in Mark alone). This is the common theme of RR heroes: the hero is an heir to a throne (vs. a usurper or new dynast, or not to be a ruler at all). There is no way to escape this parallel.

      Likewise, it does not matter in what different ways one can be a foster parent, all that matters for the peculiar feature is that one has one, unlike most people, and indeed contrary to the expectation of royal heirs and heroes generally. Moreover, scoring the point requires rearing abroad, not just fostering. And that’s a match, too. So there is no way to escape this, either. Indeed, it’s a deliberately created one, designed to match the Rank-Raglan hero Moses, by scoring this very point. The adjustments to the exact story are irrelevant to this. Likewise are “differences” with other heroes. Differences are irrelevant. Similarities score. Period. (Again, if that were to create too broad a set, the set should contain mostly actual historical persons; so that it doesn’t, is how we know the scoring attributes remain substantively distinguishing of mythical persons.)

      On the childhood matter, I discuss this in OHJ, downscoring Jesus or upscoring him depending on which texts you credit as relevant to the original Gospel model (like the Infancy Gospels, which I specifically there mention). We likewise have texts that give him more criteria than the canonical Gospels (e.g. other Gospels made his parents relatives of each other) and one can even upscore him on abstracted criteria (e.g. Jesus actually does marry the daughter of his predecessor: he is identified in the Gospels as the groom, the Church the bride, the daughter of Old Israel, thus the twelve tribal fathers replaced by the twelve disciples). So the various possibilities are covered in my treatment already.

      That Jesus was not acknowledged the king by everyone is not relevant either. He was so acknowledged by the people (in the triumphal entry, explicitly quoting and citing and replicating the events of Zechariah 9, which is about hailing the king). And even ironically by the state (in his crucifixion inscription, and Jesus tricking Pilate into admitting he was the king of the Jews: a joke the authors put there on purpose). The Gospels very clearly struggle to establish this element, while maintaining the verisimilitude of not inventing a state-recognized king history did not record. So it very definitely scores. It is indeed a very good example of how hard the authors deliberately tried to make Jesus conform to the trope, while still fitting the actual recorded history of Judea. That’s evidence of mythologization, the very thing this trope signifies.

      I rewrote the “wars and natural catastrophes” criterion (as I pointed out I did others, too) to match the actual facts of the Rank and Raglan lists. Their wording did not match how they were scoring, as Dundes demonstrates. The common attribute was never “reign is a complete blank” (almost no one on the list scores that). It was always “reign lacks wars and catastrophes” (and not even all members score that, though Rank and Raglan did correctly downscore accordingly, as do I). Again, what most of the RR heroes have in common is the absence of conquests, invasions, and national emergencies; contrary to the expectation of heroic and other mythologized but real kings (like Alexander the Great). That it scores is thus exactly what is telling. If you were concerned that opening this criterion to include more people would have the effect of roping into the set more real historical persons (the only complaint you could possibly have to broadening the criterion), you should then observe it having done that. That it didn’t, is precisely what establishes the criterion as indicative of mythologization, when found in conjunction with a majority of the other criteria.

      Likewise with prescribing laws, Raglan is not Moses. What he wrote is not Scripture. What scores in my set is any declaring of laws for one’s people. Jesus did that. Getting hyper-specific about legislative procedure makes no sense, particularly when we are looking for mythical figures, not political history. The Gospel of Matthew especially went out of its way to establish Jesus as the new law-maker of the Jews, by making him the new Moses, with the new Pentateuch (the five books of Moses replaced by the five discourses of Jesus on the law), revising the Torah and replacing it with a new set of dictates. This is indeed exactly what’s peculiar about Rank Raglan heroes in respect to their establishing new laws, and in being a lawmaker.

      And yes, Jesus is literally driven from the capital city. The very city whose populace hailed him their king with hosannas and quotations of scripture just days before. He is tried in the palace of the high priest and the praetorium in the city, then the people unexpectedly reject him en masse, calling for him to be crucified, and in result he is dragged out of the city to be killed on a hill. That merits.

      Likewise, getting hyper-specific about what constitutes the holiness of a sepulcher does not match the facts of the trope. Most of those who score, score for still having tombs despite their bodies having vanished, not for the tombs being shrines. Again, making this broader, should bring more historical persons into the list, not less. So there can’t be any principled objection to broadening the criterion to include all entombed persons. Whatever prior odds of ahistoricity result, can then only be lower, not higher.

      And the contest with Satan is certainly an emulation of solving the Riddle of the Sphinx, the paradigm model used in the Rank-Raglan scheme. It very definitely rates. The differences are irrelevant. Satan is an adversary, and as Satan himself proposes in the encounter, a rival king. Satan was well known to the composers of the Gospels as the dragon (of Eden fame), even a rival superpower (as Paul said, Satan was the God of This Age). But clearly even in the Gospels, a rival ruler, whom Jesus must defeat before the triumphal entry. A sequence that peculiarly matches the ubiquitous hero trope. That he doesn’t use a sword is irrelevant. Just as Oedipus defeated the Sphinx (and thus scores) by word and will alone, so does Jesus.

      Oedipus is also completely uninterested in being a king and is later surprised to find out he is to become one; so again, clearly that has nothing to do with the criterion, and is an irrelevant difference among the manifestations of the same trope. You need to realize that these are the same bogus techniques used by Marshall: claiming different ways to realize a trope, eliminates the trope. False. The similarities establish the trope. The differences are what keep every hero on the list from just being Oedipus. If such trivial differences mattered, there could never even in principle be fifteen members of the set, because only the same one identical person could be in the set!

    2. Sorry, I utterly fail to be convinced either by your method or your arguments.

      First, while you obviously do not need to follow Rank or Raglan, and you can indeed invent your own scoring, this opens up a massive risk: that you are in fact unknowingly engineering this classification specifically for the purpose of arriving at a preconceived conclusion. And that appears to be very much what you are doing.

      Second, it should be noted that this method is, whether we follow an established pattern or create a new one, entirely spurious. What would be the causal connection between fitting such a pattern and being or not being historical? It is on par with sports pundits citing meaningless statistics such as the Tigers always beating the Walrusses if it was raining the Tuesday before the game.

      Like in that case, there are an endless number of possible lists we can put Jesus in. Especially so when we are allowed to invent our own lists and fudge the criteria in order to fit Jesus. The supposed low probability of Jesus occurring on a list with a number of non-historical people is entirely offset by the astronomical number of possible lists that we could choose from.

      Then there is again the way you interpret the stories in order to make Jesus fit. For instance, on Jesus being a king (rather than being expected to become a king at some time after the story has ended), you claim this is evident by the story of Jesus’ triumphant entry into Jerusalem. But, Mt 21:10-11:
      »When He had entered Jerusalem, all the city was stirred, saying, “Who is this?” And the crowds were saying, “This is the prophet Jesus, from Nazareth in Galilee.”«

      Jesus, in the Gospels, is not a king. He has no territory and no subjects. He has no Earthly power at all beyond his twelve not very loyal followers. He is not referred to as presently being a king, although he is expected or prophesized to become one at some point in the future, not covered by the story. Pilate does not ‘admit’ Jesus is king, he is mocking Jesus as obviously not being the king people expected him to become (Pilate of course not knowing that Jesus will rise from the dead and reign for 1000 years any day now, the fool that he is). Jesus in the Gospels is a preacher and a prophet. That is what he does, and he is referred to as such. And this is why his story quite frankly is a poor match with those examined by Rank and Raglan, despite a few striking similarities such as the murder attempt at infancy, being the reputed son of a god, and losing favour with his supporters. But then, causally and logically, why would an historical, highly mythologized person not be attributed with such stories? If the story is good, then it will be reused.

    3. You need to attend to the mathematical reality: “that you are in fact unknowingly engineering this classification specifically for the purpose of arriving at a preconceived conclusion” is impossible at the resolution I use. You cannot “accidentally” get 15 members and all nonhistorical by any random or even designed list of criteria, unless there is a statistically significant relationship between the criteria and nonhistoricity. There can be a margin of error, but mine is so wide it more than amply includes it.

      As to the rest (causal connection etc.), that’s all addressed in my book, even explicitly in named sections (Ch. 6.4, “The Causal Objection”). Correlation is correlation. Causes don’t matter. Unless you can show a cause applies that changes the probability. So, you are welcome to find data that shows that it changes something. That’s how this works. Likewise, if my interpretation of criteria brings in more historical persons than I claim, you have to show that. If it doesn’t, you can’t complain about my interpretation: because it still evidently only draws in mythical persons!

      You have to stop being obsessed with the RR class and with treating Rank and Raglan like scripture. You need to start paying attention to how the math entails the conclusion no matter what criteria are used or how they are interpreted (as long as it’s consistently). You need to start making arguments from evidence and not your unevidenced feelings.

      Overall, it’s always going to be a fact that the more mythologized a person is, the more likely they are to be mythical. The set of “very mythologized persons” is obviously going to contain a higher frequency of mythical persons than the set of “not very mythologized persons.” If you want to show the frequency is different than I adduce for persons as highly mythologized as Jesus, you need to get data that shows that to be the case. You can’t just sit in the armchair and declare the frequency to be different. I have data. If you want to show the frequency is different, you need data. Where is your data? Where is your set of all the persons in history as mythologized as Jesus? How many are in that set? How frequently are persons that mythologized historical?

      Answering that is the only way you can updaate my estimate that the answer will never be higher than 1 in 3. Obsessing over irrelevancies and ignoring the mathematical realities of the RR class will not get you anywhere factually credible. You need an actual set of persons as mythologized as Jesus. It needs to have at least ten members to give you a statistically significant frequency. Go find that. Then we’ll talk. Until then, I’ve gathered far more statistically significant evidence than you; so I’m working with knowledge. You are not.

    4. I’m aware that you think causality does not matter. I don’t think many others are convinced of this. From a very strictly logical view you are right, given certain assumptions. However, these assumptions include that we can be sure that our data is independent, that we have not knowingly or unknowingly filtered our data to fit the hypotheses, that the criteria are the only possible criteria, and that the mathematical model in other ways fairly represents the real world methods used to end up with the numbers used to calculate a probability. There is, rightly, a lot of skepticism that all such conditions are fulfilled, especially in a case with such a small data set.

      Compare with a medical clinical trial: here, even if we have tested a thousand patients and 60% claim that their nose bleeding got better after they started eating popcorn every day, we will still rightly be skeptical that popcorn really has this effect. Even if we can’t find a flaw in the study. Rather, such a finding will set off research into possible mechanisms for how popcorn could effect nose bleeding. If the trial sizes are immense and repeated, we may attach some confidence to the results. But we will still not be satisfied until we find the causation.

      I think what you have done reminds quite a lot of the social scientists that invariably pop up in years when there is a US presidential election. They always claim that they have a simple model that based on just a few parameters will accurately predict the result of the election. And they show that their model indeed works for the past 15 elections or so.

      Now, these models are not actually reliable. Unlike with the sports pundits, or your criteria, the parameters (usually) have a viable causality. For instance, if the economy according to some metric gets worse, the incumbents’ chances of winning are reduced. But what is really happening is that the scientists are engaging in overfitting: taking a very small number of events (US elections in the modern era, or mythical ancient hero-kings) and laboriously finding a way to get to a desired outcome (the model predicts the winner, or Jesus fits the group).

      It is simply not true that there is such a limited ‘margin of error’ as you think. It is very easy to design criteria that will make, for instance, any particular US president fit in with this group of ancient heroes. Or any Nobel Prize winner, or really any person for which we have enough data. I know that in OTH you argue that you are not engaging in this because your criteria are ‘meaningful’, but although some criteria can convincingly be argued to be irrelevant (‘number of letters in the first name is equal’), we must expect a very wide range of opinions and interpretations on this point.

      The problem is not, as you seem to think, that we may find historical persons that fit your criteria. The problem is that there is a practically unlimited set of criteria that you could have used. Most of these criteria will either not fit Jesus or allow (many) historical persons to fit the group. You chose the ones that gave the desired outcome. It is similar to what I call the TV lottery fallacy: we see the number 32902992 on TV and it turns out to be the winning number. What are the odds of this happening?! Slim to none, it seems. But of course not. The number was chosen from all possible numbers precisely because it was the winning number. We should no more be amazed that your parameters, chosen from the billions of combinations of parameters you could have used, give a small probability of an historical Jesus fitting them, than we should be amazed at seeing a winning lottery number on TV.

      So I’m not at all ‘obsessed’ with RR. I’m not at all convinced by Raglan’s theory of myth originating as rituals and his scale is indeed old and obscure, although the parallels drawn are, at least in part, interesting. But, at least, we can reasonably assume that Raglan did not design his parameters to fit Jesus.

      As for highly mythologized persons being less likely to exist, this seems a bit tautological because non-mythologized persons should by definition have existed. But at any rate this is a separate issue than the argument from RR-esque criteria.

      I don’t think we will be getting much further, I will only conclude by opining that for me this argument reduces the credibility of the mythicist position, much in the same way as Price’s “everything is a forgery” stance. Whether rightly or wrongly, it immediately strikes me as fundamentally unsound.

    5. “It is very easy to design criteria that will make, for instance, any particular US president fit in with this group of ancient heroes.” You are not paying attention to the math then.

      First, not even one President fits these criteria. Much less fifteen of them. That’s the problem. Not with whether one person can be twisted in. But how you can imagine it likely we can twist in six or ten, much less fifteen. You do not seem to understand why that is extraordinarily improbable.

      Second, if you can get historical people in (you claim to be able to, but have never demonstrated it), then that will raise the frequency of historical persons in the set. However, if there is another set to which Jesus belongs in which all other members are mythical, you must apply that instead.

      So, for example, if we interpret every criterion on the RR scale so widely that half of all people fit, then you’ll get a set in which most members are historical. But if a narrower interpretation of those same criteria reduces the number of historical people in the set, yet keeps Jesus in the set, that means Jesus also belongs to a set of low historicity members. And if, after doing that, there are still over ten other members of the set, the correlation is too statistically significant to be accident or design. Such a high membership is very improbable, unless the set is real, and the correlation predictive (of Jesus as much as every other member of the set).

      By analogy, I can have criteria for detecting cholera, such that a wide interpretation gives me a high false positive rate, but a narrow interpretation gives me a low false positive rate. Obviously it is invalid to prefer the broader to the narrower interpretation. Conversely, I can have criteria so narrow they give a high false negative rate. Obviously I don’t want that either. So the criteria must be narrow enough to minimize false positives but broad enough to minimize false negatives. Moreover, the criteria must be broad enough that there are enough members of the resulting set to give a statistically significant frequency by; and as narrow as possible within that condition, to minimize false positives.

      Likewise with RR: we are looking for cholera (probability of having cholera = probability of being mythical); we have a set of criteria interpreted narrowly but still broadly enough that there are over ten members in the set; and we confirm every member in that set has cholera; leaving us, at most, a 1 in 3 chance anyone else passing that test actually doesn’t have cholera but just looks like they do (= is actually historical but just looks like all those other mythical persons). You can’t gainsay this by saying “but if we interpret the test criteria more broadly, a bunch more people who don’t have cholera will test as having it!” That would simply be a faulty test. Making the test less accurate at detecting cholera is not a way to claim the test we are using is invalid!

      Likewise you can’t gainsay this by saying, “But maybe you designed the criteria to fit those ten plus persons who have cholera so the result doesn’t predict the odds for other persons meeting those criteria!” Because that would be statistically impossible for so many persons. If over ten persons both have cholera and meet those criteria, and no one without cholera meets those same criteria at that same narrowness of interpretation, then that is exactly what we are looking for to detect cholera! The odds that a new person would both just “accidentally” meet all those same criteria and not have cholera are necessarily low. No better than 1 in 3, for example. See how this works? You can’t escape this with maybes. The math can’t be tricked.

      In the end your words are just wasted breath without data. That there “might” be a problem here somewhere is irrelevant. You have to show, with evidence, that there is. You can’t say “maybe some causal factor changes the correlation frequency.” You have to show with evidence that it does. I am arguing from evidence. You are assiduously avoiding ever basing any of your assertions on any evidence that any of your suspicions are true in this case. Only one of those methods is properly rational. If you ever have any evidence you want to talk about, I’m game. But your endless series of maybes doesn’t change the evident frequencies. So it’s all idle discussion. Until you present evidence any of your concerns actually change the frequency here. Not “might” do so. Actually do so.

      If you want to gainsay the conclusion “at best only 1 in 3 persons who get as mythologized as Jesus really existed,” you need to show with evidence that there is a set of persons as mythologized as Jesus in which more than 1 in 3 are historical. And that set needs to be large enough for the frequency you find in it to be statistically significant (probably at least ten members, certainly more than six). Everything else is just armchair speculation worth less than a wish and a nickel. Get evidence. Then we’ll talk. Until then, I’ve got the evidence, you don’t. That wins, in every empirical field.

    6. Johan you do make good points and a lot of what you say is valid. But you might be missing the forest for the trees. The gist of the argument is that Jesus’s biography is so mythologized to the extent that the only people who equal that level of mythologizing are nonhistorical, and that no historical person has ever been mythologized that much. This premise holds true independently of whether the Rank Raglan criteria are valid or whether Richard modified them, or even whether a whole new set of criteria is used in the first place. Either way, Jesus will end up in the company of overwhelmingly nonhistorical characters as opposed to historical characters. This is an important fact that has to be factored into the probability calculations somewhere and cannot be ignored or brushed aside. The real question is whether there is any reason to believe that more than 1/3rd of such mythologized characters are based on historical people. And if so, how much more? It would have to be a lot more to challenge the final conclusion of OHJ. Even a doubling of the prior probability would still leave a final probability of historicity of only 66%, which is hardly a solid foundation to justify the bravado and bombastic confidence of the historicist camp.

  3. Denis Gaudreau January 22, 2017, 3:14 pm

    Hello Dr Carrier, good topic again ! Well I’m agnostic, born and raised from french canadian catholic background. My view over that complicated topic did Jesus ever exist or was only a Jewish myth, I do look at it on a practical and factual terms now. My question is : is it possible to a human being to have some of the supernautral features Jesus had ?? Since the Stan Lee Superhumans show aired on TV or that I’ve heard about that Netherland’s man named Wim Hof that can stand incredible freezing weather, my opinion is now that Jesus might have been a preacher / Jewish prophet of his own rooted with John the Baptist and perhaps the Essenes. Also as I said I was born french canadian catholic, so my mother used to told us tales about a catholic friar named Brother André / Andrew that died in 1937 and was broadly known not only in the Province of Quebec, but from Canada and Northern US States as well. He was known since the early XXe century as the Miracle man, performing miracles and healing that was enraging the English physicians and doctors from Montreal and suburbs as they were losing clients to him. He was never showed as a fraud and he was performing his deeds in Joseph’s name, father of Jesus. Probably that friar had some of these “superhuman” gifts like ESP, telekinesis or else. That’s why I can think and accept that there was a man Jesus doing something surnatural to some extends but after his death, everyone tried to have him on his side stating the he was from Jewish king lineage, or the Messiah or Son of God. Gospels are then nothing more but boastings and exagerating deeds of probably a real human that had some surnatural gifts. To me it would make more sense, why non Jewish people like Paul and others get along with that story. Otherwise it is pure non sense to have so many people from different cultural backgrounds to leave behinds famial beliefs for a Jewish ghost or angel, whatsoever. Even now I wouldn’t be surprised if Hercule and others antique hero were just humans with superhuman capabilities as well.

  4. Peter January 23, 2017, 7:51 pm

    Is it even a surprise anymore that Christian apologists are pathological liars? They’ve been lying since the beginning of the faith with all the forgeries, redactions, and interpolations in even their own scriptures.

  5. David Quigley January 23, 2017, 9:18 pm

    First off great article and podcast. I agree that Unbelievable is usually a fair and good discussion of topics, as is demonstrated by your previous two appearances. However, I feel that this one was a little less fair, not because you had to deal with a liar who is very good at avoiding points and questions, but because he not only got the first word, but also the final word. I think this could have been handled differently. Was it edited weird, or was this the actual case?

    Second, I have a quick question, and you may have covered it in one of your books, but I missed it. You mentioned in this entry that “…Jesus is one of the most mythologized characters in the whole of human history” if we add the RR Class to his already staggering mythic qualities. That being said, do you think that Jesus was so popular precisely because he had all these traits added on him? It is strange that Marshall never considers that Jesus has all these traits on purpose because they were added to make him the end-all amalgam of all the mythic characters of the time. Am I completely off on this? Thank you!

  6. Update: Many people found my addition of hyperbolic emotional language, that such a liar is a disgusting person “who can go die in a fire,” too offensive, so I removed the quoted phrase, both to prevent anyone from mistaking me for meaning that literally, and to remove the distraction of my emotional expression of outrage, so people can pay attention to the facts—and hopefully cultivate some outrage of their own, at the one who actually deserves it.


Add a Comment

I only publish comments by my patrons and anyone who or whose work I discuss in the article commented on. Comments must also follow good etiquette. Those who engage in dishonest, abusive, or harassing behavior may even be banned as commenters and patrons.