In my recent update on the “Jesus Myth” debate, in The Obsolete Paradigm of a Historical Jesus, I note that, now, the best reason to be sure Jesus didn’t exist is that everyone who insists he did has to strangely ignore a lot of evidence (even pathologically avoid it), or else make up fallacious arguments based on no evidence or wild distortions of it; or even sometimes outright lie. Any conclusion that was surely true should have had a real defense by now. Not that. Ten years, and not a single good argument. I made this point again just last month with the hilariously terrible case of Ken Dark on the Historicity of Jesus.

I maintain a List of Responses to Defenders of the Historicity of Jesus. You can check every single one. It’s the same story. Every time. Maybe not as hilariously bad as Dark. But generally, bad. In How to Successfully Argue Jesus Existed (or Anything Else in the World) I tried to correct the common mistakes pervading these attempts to defend the historicity of Jesus, in a broad sense (things I honestly shouldn’t have to explain to experts yet evidently do, like, “Tell the truth,” and “Be your own best critic”). And I’ve listed all the usual worn out tropes in Historicity Big and Small: How Historians Try to Rescue Jesus and its sequel A Few More Attempts to Rescue Jesus. And I’ve catalogued countless examples of how Things Fall Apart Only When You Check: The Main Reason the Historicity of Jesus Continues to Be Believed (which I updated and matched with scholarship and evidence in Obsolete Paradigm).

But most scholars avoid this debate altogether—often while continuing to move the chessboard in my direction anyway. I give numerous examples in Obsolete Paradigm, but maintain An Ongoing List of Updates to the Arguments and Evidence in On the Historicity of Jesus that reflects the same trend. Gosh. It used to be there were no dying and rising gods, we made that up, or no Jews would ever believe in a dying messiah, only amateurs would overlook that. Now it’s dozens of mainstream studies proving there were and they did. And so on. The rising mainstream trend is now toward the Gospels as myth, not records of oral lore. The number of bona fide experts taking doubting the historicity of Jesus seriously is now approaching fifty (see my ongoing List of Historians Who Take Mythicism Seriously). And so on. But nevertheless, the third rail mustn’t be touched.

I’m starting to think the only reason any honest scholar says they are “confident” Jesus existed anymore is because they believe they have to. That belief might be false. But it doesn’t matter. How things seem becomes how they are. If a scholar merely believes that admitting doubt will produce endless headaches of pressuring, meddling, harassment, and labor (or damn their soul, or embarrass them before their fans, family, peers, or employers), they will be well enough motivated to simply avoid the entire question of historicity if they can, and give lip service to it if they must. While the most fanatical or narcissistic will pump their chest and browbeat to secure their reputation and social status from stereotype threat (these tend to be the liars of the bunch).

But that leaves an undiscussed circle on the Venn Diagram: historians who take doubt seriously, but still favor historicity. These scholars will admit that, yeah, it’s entirely reasonable to doubt Jesus existed, and that scholars do need to take that doubt seriously and not pretend it’s fringe. They have no problem admitting mythicism is plausible and could be correct. They’ll take the heat for saying so out loud. They just suspect that, yeah, probably there was nevertheless a guy. We can’t know anything about him. And almost everything the field says they think they know about him is dubious or false. But, still. On balance, probably there was a guy. Can’t be certain. To claim certainty in the face of the actual scarce and compromised evidence we have is what’s actually embarrassing. But, you know, on balance. After all, there are perfectly plausible models of a historical Jesus—as even I have described myself (in a fun talk in Wichita years ago, and since on the MythVision podcast).

The Last Three Arguments

I think overall there are two kinds of scholars still in that camp: those who aren’t aware of the state of the field on some fact or other, and are thus just one lit review away from doubting more; and those who are totally up on the studies and relevant background knowledge, and are instead hung up on some point of conceptual logic.

• The first camp includes, for example, people who are still psychologically hung up on the “someone would have said something” argument (P34 in my Few More Attempts), not being aware that the field has actually moved on to basically admitting they did. Likewise those who think the celestial myth model is ad hoc—when in fact it is thoroughly precedented in adjacent Osiris cult (OPH, 23–24) and even Jewish literature (OPH, 73–74) and beyond (OPH, 14–15, 41–42; OHJ, Elements 34–42; see also How Did Christianity Switch to a Historical Jesus?).
• While the second camp includes, for example, people who agree (I’m here paraphrasing a colleague) that “there is no real attempt at preserving a memory of a historical man named Jesus anywhere in the New Testament” but “rather, the opposite,” indeed, “to an extreme,” but who still have conceptual problems of logic holding them back from going all-in on doubt—which are not problems of being uninformed, but of navigating some of the complex rational relations among ideas.

Of those latter hang-ups I’ve heard three so far (again paraphrasing):

1. We don’t really have any strong conventional pattern for whole-cloth invented-yet-historicized figures through that period who were portrayed in such proximity (in time and geography) to their alleged existence. Yet we have many mythologized historical figures. Jesus would just be an extreme example.
2. The prosecution and execution of a hero by the state seems like too much of a morally compromising stain for an origin story. The texts we have look intended to paper over and reinterpret that stain.
3. The “holy family” (with Jesus and his brothers as cult leaders) fits as a power-locus in the origins of religions generally. The Sunni/Shia divide in Islam, as an analogue, came down to that conflict early in its history. Christian sources seem to evince a debate over whether the family of Jesus are the true tradents of the cult, or to be dismissed as “in the way.” Which seems to suggest a historical Jesus.

These have one colleague of mine “leaning” toward “there seems to have been a guy,” but “not with anything approaching dogmatism,” in fact not seeing this “as even all that important to the field.” In the same way that “the historical Saint Nicholas need hardly even merit mention in a full cultural study of Santa Claus in Western society.” The real Jesus is unknowable anyway. So why should we even care whether he existed? We wouldn’t need to care about Santa Claus existing. It’s not as if discovering there wasn’t even a Saint Nicholas would have any effect at all on modern Santa mythos or its wide embrace across society. Scholars of Santa Claus barely even need discuss Saint Nicholas, and if you cut the few pages they spend on it from all their studies, nothing really would change about the rest.

The Logico-Mathematical Effect

The key to understanding all three of these objections is to try and articulate what effect the cited data would have on the probabilities. They are moot if the effect is none. But the only way they can have any effect, logically, is by collapsing the prior or inverting a likelihood. Which in effect reduce to the same thing, since one way to test the effect of a point is letting it have its effect as a likelihood to update the prior. I run through examples of this process in Chapter 6 of On the Historicity of Jesus. Because every prior is technically the output of a previous run of likelihoods. Even if you start with a completely uninformed prior, that, too, is developed directly from a “lack” of any evidence, background or otherwise, weighing for or against the hypothesis (per The Principle of Indifference).

Here we have an informed prior, based on the frequency of heavily mythologized superheroes being historical (which on my upper margin updated now in Obsolete Paradigm is 1 in 4). But do any of points 1, 2, or 3 change that? On my upper margin I find the net effect of all evidence to be basically a wash (thus making little substantive change to the prior), because even when counting a bundle of evidence 8/1 more likely on historicity, the rest of the evidence is in total nearly 8/1 more likely on myth, so those cancel out. The only way these three new points can change this outcome is to establish that their required premises (major and minor) are all true and are more likely on historicity than on myth—and then estimate how much more likely. And here, since we are seeking the upper margin (the best odds on Jesus existing), we would ask, what’s the most likely it is reasonable to believe. If anything more would be unreasonable, we have to stop there, as then our resulting conclusion is also unreasonable.

So really, in all three cases, what we are asking is: how likely are those facts (if they are true) given the least ambitious historicist theory (just bare historicity being asserted); and how likely are those same facts given the least ambitious mythicist theory (the one requiring no assumptions about the state of things then that are not backed by at least some evidence, like the background evidence surveyed in Chapters 4, 5, and 7 in On the Historicity of Jesus). For more on all this, see Doing the Math: Historicity of Jesus Edition (and, for an example of quite bad “final arguments” for historicity run being down mathematically, see How to Correctly Employ Bayesian Probabilities). For more on the logic here, see Chapters 5 and 6 of Obsolete Paradigm.

Does Proximity Matter?

In a statement like “We don’t really have any strong pattern” of examples for an invented hero “in such proximity (in time and geography) to their alleged existence,” all the heavy lifting is being done by the word “proximity.” If you remove that clause, the point becomes false: we have a very strong pattern of examples of historicized heroes, and Jesus matches it exactly (Chapter 6 of OHJ); in fact, Jesus matches more mythic patterns than historical ones (Obsolete Paradigm, 225–46). So, really, Objection 1 is simply the claim that it is improbable that this would be done so quickly—whereas the data then are more expected if this was built on top of a real guy. I already explained in Chapter 6.7 of OHJ, “Rapid Legendary Development,” why this likelihood differential does not result as expected. But I expand on the point several different ways in OPH (“Not Understanding the Alternative,” 140–43; and 15–16, 119–120, and 220–25), and all the way back to my resurrection debates, when it came up for that (when I already had to explain that, No, Mr. Christian, A.N. Sherwin-White Didn’t Say That. And Even What He Did Say Was Wrong.)

But why this is the case is not obvious, so it is easy to fall into the trap of this reasoning. “Why has this never happened before?” is an instinctive question to ask. But it’s flawed from both sides. First, we don’t know that. How soon after people were invented were they historicized? In some cases we know (it’s hundreds of years for Romulus and Moses); others, not so much (Dionysus? Hercules? Aesop? Pelops? Oedipus?). And although on balance I think Apollonius of Tyana existed, only to a marginal probability, it is not because of proximity that I think that (OPH, 220–25). So we can’t really say. But we do have precedents: not only do I cite Roswell, John Frum, and Ned Ludd (which should have been far harder to rapidly invent, given far more difficult literacy and media environments to survive in: OHJ, index), but I also cite the Risen Jesus.

As I put it in OPH (15–16):

Objective historians today must admit there was no resurrected Jesus—no animated corpse rose from the grave to hang out and have dinner with his followers. That man is not historical. Yet he became so, within a mere generation or two. And the original, mythical version of him (who was entirely imaginary, as Paul reveals in Galatians 1:11–16 and 1 Corinthians 15:3–8) was erased. But if a ‘historical’ Risen Jesus could be invented, and that version narratively erase the original (the one of mere fleeting dreams and visions), then so could a ‘historical’ Jesus prior to that be invented, and that version narratively erase the original. We cannot allow the one sequence of events to have happened (as plainly it did), while denying the other could. The question thus falls to the evidence: does the preponderance of evidence argue more for the traditional account, or this other?

In other words, we can’t argue “no one else invented a resurrected hero so quickly, therefore Jesus was resurrected” (or his resurrection was pretended by some elaborate scheme or whatever). That simply does not increase the probability of that at all. That this rapid development happened (or happened to survive in our documentation) only once has no effect on the probability of its being a myth, any more than it affects Roswell (from scattered cheap junk to whole spaceships and autopsied pilots) or John Frum (from whispering telephone poles to a whole national hero) or Ned Ludd (from a diffuse anti-industrial zeitgeist to its suddenly having a made-up founder). This is obvious in the Risen Jesus case (the original visionary event was replaced by cuddly dinnerparty Jesus in just forty years time). Yet it should be just as obvious for the premortem Jesus as the postmortem Jesus. Both were invented on the same timeline by the same people at the same time. So obviously there was no obstacle against it. This one being documented quickly is just an accident of circumstance, one no more improbable whether Jesus existed or not.

So we cannot assign either of these outcomes different probabilities. It is no argument that Jesus “must” have survived the crucifixion and pretended he rose from the dead because “we don’t really have any strong pattern” of dinner guests being so rapidly and elaborately invented after their actual deaths. “It was too soon for an empty tomb to have been made up!” is not a plausible argument in apologetics. Yet it’s the same argument as “It was too soon for a preacher to have been made up!” The background evidence shows that it is not too soon. In fact, Roswell, Frum, Ludd, and the Risen Jesus all evince the fact that forty years is a typical pace of invention. So it is not improbable. And if it’s not improbable, we get no inverted likelihood ratio. The probability Jesus existed is logically unaffected by the “proximity” of invention—unless the proximity is days or years rather than decades, which is not the case here. Here, “proximate” means a whole average human lifetime later. That’s actually not very proximate. It is, rather, the start of the “time enough for it” window, as again the proof cases show (Ludd, Roswell, etc.).

The question might then be raised “should we have examples from antiquity, rather than modernity,” given that our best proofs of concept are mostly modern (Roswell, Frum, Ludd). But this is a product of happenstance and evidence-survival. The trend for all examples (all historicized myths of whatever kind, in all eras of history) is to “place” invented persons in history at the believed “origin” of whatever their myth is supposed to explain or be a part of explaining (OHJ, 249). Hence “Moses” and “Osiris” get placed when it is believed their cults began (or when their inventors wanted people to believe they began). Romulus, at the believed founding of Rome. Most Greek heroes get piled on to when it was believed the very Panhellenism they represent began: with the Trojan War. Aesop appears to have been placed in history when it was believed his fables first appeared. Homer, when it was believed the Iliad and Odyssey were written—though in fact that was the period of their final edit; the books were written across so many centuries their narrative spans the bronze and iron ages, as their heroes’ weapons and armor change according to the era a given scene was written in. Thus, “when” heroes get placed in history was always a function of what they explained, not how long ago it was.

Even Jesus was placed in different times depending on when someone thought the religion began that he was invented to found. Some Christians outside the Roman Empire placed him in the Hasmonean era a hundred years before the Roman era—and these were Torah-observant Christians who would not have heeded any writings of Paul, allowing more leeway as to when they could pretend their religion began (and we know they calculated that from scriptural math rather than historical evidence: OPH, 57–60; OHJ, 281–88). But Christians inside the Roman Empire, following the writings of Paul, saw (or, as they were under the Roman empire, believed) the religion began in the 30s, and therefore had to place Jesus there. The proximity (or lack thereof) was irrelevant: it was decided by the happenstance of already-existing beliefs about when the thing they mythically began “began,” not by any need to wait longer.

Since the timeline is therefore a product of factors other than the need to wait, and those factors were completely random (which century the thing to be explained was believed to have started, or needed to have started), we cannot expect many examples of short timelines. Picking centuries at random will give you one or two short runs but mostly long ones. Greek heroes were stacked into the 12th century B.C. because that’s when the Trojan War was believed to have happened, and all Panhellenist myths therefore had to be set there. This was a complete accident of circumstances; not a need for a long time to have passed. So that they all get stacked “there” and not “last century” is fully expected even if “they needed time” was false. Therefore “they needed time” is not established by any evidence (or at least cannot be established on this evidence; it is refuted by the modern examples and the Risen Jesus example).

Since we cannot expect long or short times, because the timeline is accidental, and since the timeline for Jesus was fixed by when they thought the religion began (and indeed even Paul is compatible with both chronologies) and not by how recent or not it was, it is no longer the case that we should “expect” a longer timeline for the Jesus myth. Therefore we cannot argue that it is “improbable” that the timeline would be short. We also can’t argue we should “expect” many more short timeline heroes, since all the heroes we know had their timeline fixed already by other factors, not recency or latency. So the absence of other examples is not improbable either—it’s just as likely on either theory. So we get no inverted likelihood ratio. The “proximity” has no effect on the probability of historicity or myth.

But really, we do have a “strong conventional pattern” for rapidly invented people then, because so many people were fabricated in Christian lore on the exact same timeline: Mary Magdalene, Bartimaeus, Nicodemus, Simon Magus, Judas, Stephen, Barabbas, Joseph of Arimathea, Elymas Barjesus. That itself already refutes the contention that it was too soon to invent people (this continues with the rapid invention of many Christian saints through the middle ages). So “we don’t really have any strong pattern” of examples for an invented hero “in such proximity (in time and geography) to their alleged existence” is simply not improbable on mythicism; ditto for even more famous events invented, not just people, from alien-star sky-magic to blotting out the sun or single-handedly going Bruce Lee on a heavily guarded temple square—and a great deal beyond.

So early invention of a myth cannot change its probability as determined by other evidence. It only could have had the myth arisen mere years later, not decades (see OHJ, 54–55). But that is not the factual situation here. Here, all precedents, ancient and modern, establish forty years to be ample time for the wholesale invention of any person or event. And circumstances required this for the Christians to exploit the Jewish War: that event essentially necessitated creating a historical Jesus to sell their apologetic mission at that point (the same reason that they had to have only gotten around to this then, and not earlier or later, even if Jesus did exist); and Jesus had to be placed only forty years before that, because that’s when the cult began that his myth was meant to explain (or so the authors of our Gospels believed—or wanted people to believe: OHJ, Chapter 8.1). So contrary to intuition, the probability of this outcome is the same on myth as on history.

I go more into the mathematical logic of all this in Jesus and the Problem of the Fraudulent Reference Class, where I explain why (although there in the face of ridiculous rather than plausible attempts at similar arguments) we mustn’t replace a prior probability already developed on a known and relevant reference class with a less known or relevant one. We must heed the data as it is, and its actual logical relation to a theory. And here, the proximity of a whole lifetime later has no logical relevance in the face of the actual data we have on how origin myths get placed in history.

Does a Stain of Criminality Matter?

Second is basically what I years ago labeled Argument P15, “The idea of a messiah who would become victorious by dying, would never occur to a Jew, unless confronted by an actual candidate (like Jesus) being killed.” Which, normally, would be a worthwhile point: absent any other data, it would be unlikely this would be invented as a myth, rather than a historical problem that had to be explained away with a myth. So again, the reasoning superficially makes sense. But we actually have evidence this case was an exception: unexpectedly, we have a ton of evidence that the creators of this religion (who were counter-cultural apocalyptic Jews) not only expected but needed a hero to be criminally convicted and humiliated and brought to the lowest of the low. When we bring this background knowledge in, it completely changes our expectancy. Now this being invented is just as probable as it being historical.

And so, again, no inversion of likelihoods results. In fact, we’re left at a wash: this is no longer evidence for or against historicity. It is equally expected either way. I think a lot of people mistake my saying this as an epicycle, that I am just “assuming” this could be the case and “therefore” we can assume the likelihoods are equal (which would indeed be fallacious: see The Cost of Making Excuses). Rather, this is a straightforward conditional probability: rather than an assumption, this exception-premise is an emipirically proven fact (see The Idea That Some Jews Were Already Expecting a Dying Messiah although I have a peer reviewed journal study coming out in a few months on this same point; I also add more data in OPH, 70–72, 320–21). Which, being a fact, raises the likelihood of a crucified savior on mythicism all the way equal to historicity.

Indeed, this is why I think the most plausible model for a historical Jesus is the same thing: everyone was trying to create an executed messiah to start the doomsday clock (all the other “Jesus Christs” Josephus documents: OHJ, Chapter 6.5). The Christians just eliminated the messy (and often violent and vexing) problem of having to produce and rely on a real one. They could just invent one killed by demons, known only by revelation, and dodge every problem the real ones were creating (Elements 5–8 and 23–29 in OHJ). This allowed near total control of the narrative; and easier avoidance of implications of the movement being rebellious or violent. Much as inventing the humiliating and debasing mythohistory of Attis castrating himself allowed the Attis cult to control how that metaphor operated to their marketing advantage. Its being humiliating and debasing clearly had no impact on preventing them inventing it. Same goes for a humiliated and executed messiah.

So P(convict|mythicism) is not P(convict|mythicism+assumptions), which would only move the improbability of the assumptions to the prior and thus not stop the bleeding: those assumptions or the fact to be explained, either way, would drop the final probability of mythicism, just as our intuition imagines. Rather, it’s P(convict|mythicism+peculiar facts). Since those “peculiar facts” are as near to certain as makes all odds (abundant evidence establishes that this fit the context quite well), there is no resulting improbability, and the likelihood given those facts is as near to 1 as makes all odds. So, this intuition fails precisely for the exact place, time, and culture where Christianity arose: it arose exactly when a desire for this kind of hero was at its peak, and even being speculated and tried several different ways. This is contrary to any other place, time, and culture, where that might not have been the case—which is why Christianity did not arise anywhere else.

One can argue it could have arisen elsewhere, but only if it had been different than the one we have. Greece and Rome also had religious reverence for the magical power of self-sacrificing heroes (Element 43; indeed, Aesop is an example: Element 46), but had this been the primary (rather than, as it also was, secondary) basis for the Christ cult, that Christ cult would look different in many subtle particulars. For example, it would not be justifying itself by reference to the Jewish atoning suffering-hero trope (OHJ, 430–32) and easily associated scriptures, but by reference to those other Greek or Roman mythologies (e.g. OHJ, 210–12). So that Christianity was built from the start on the Jewish tropes and justifications tells us that’s where it came from, and therefore that’s where we should find (if mythicism is true) ample precedent establishing its logic. And lo, we do. In exactly the form Christianity adopted.

Inventing exactly this kind of myth was highly motivated at exactly that time (Chapters 4 and 5). So it is not unexpected, and therefore not improbable. And not being improbable, no inverted likelihood results. It’s simply, again, a wash. And here, for the reason that this oddity of Christianity is causally associated with documented oddities of Jewish counter-cultural apocalyptic messianism and suffering hero narratives. So we do not need a real precipitating event. They did not need to be forced to contrive this by anyone’s actual death. They were already primed to invent it; and indeed, an invented one could easily be expected to be more successful, lacking all the defects and baggage of a real one.

We must remember, Christianity never sold anyone a defeated criminal. In fact we have no records of such a person at all, from anyone. Paul never once has to defend Jesus or Christians against this charge (in Romans 13 he is completely unaware of this—it is not the “stumbling block” Paul has to overcome with outsiders: OHJ, 565, 572, 613–15). Which means no one was charging them with it. They always sold Christ’s death as unjustified (something God tricked Satan into) and as a victory (using known Jewish magic to defeat Satan), and as a scripturally necessary step to a successful conquest of Earth from space (exactly as anticipated at Qumran). Jesus was always a victorious messiah, even militarily. It just all had to follow the clock in Daniel 9 that always started with a messiah being executed despite being innocent (a theme already riffing on Isaiah 52–53) before the true “Prince” could return with a vengeance to destroy all enemies and resurrect the dead.

Again, as I’ve often remarked, if you asked a historian of religion then, like say Varro, just a generation or two before, “What if the Jews came up with a dying-and-rising savior-mystery cult? What would you predict it would look like?” he could have predicted every key element of Christianity himself. The god, a subordinate celestial, would be an atoning suffering servant through death and resurrection, fulfilling prophecy of the coming end times, eliminating the politico-religious role of the Jewish temple, with initiation, fictive kinship, baptism and communion, all framed into Jewish models. Consulting their oracles, he would have easily identified Daniel 9 and Isaiah 52–53 as ideal foundations for this invention. And he would agree there would be no need of any real guy to play that role, any more than any other dying and rising savior needed to be real to be invented either.

In short, in its actual context, the improbability we originally intuited vanishes. And this is one of several examples of this revelation I came to after I initiated my study in 2008. By 2014 I realized all my intuitions were wrong. They all collapse upon contact with the evidence. And another case on that very point…

Does the Jesus Dynasty Matter?

Finally, the argument that “Christian sources seem to evince a debate over whether the family of Jesus are the true tradents of the cult, or to be dismissed” as “in the way” is another example of a superficially sound question: absent other data, it would be improbable for Jesus not to exist if there were good evidence his family struggled to control the cult he founded. Indeed, as I point out in Chapter 10 of Obsolete Paradigm, if Paul had ever made clear he meant a biological brother of Jesus was holding authority in the movement, that would decisively prove historicity, and I would have gotten a very different result in my original study. Of course, the problem was (and remains) that Paul never gives us any clear evidence of that—in fact, all the evidence he does give us sooner implies he had no knowledge of biological kin of Jesus being around at all. He only ever discusses cultic brethren.

So the question is: if we don’t have any usable evidence for this in Paul (nothing there is improbable on mythicism given our background data on how Paul used kinship language), then why do we even believe there was a Jesus dynasty? Is it even true that anyone had such a notion? The idea that “Christian sources seem to evince a debate” over it refers to very late sources entirely. The first time we ever hear of any brother of Jesus having any authority in the church at all is in absurdist apocrypha related by Hegesippus (which even he may have been embellishing into dynastic references, not the lost apocryphal Acts of James he was quoting: see OHJ, 326–31; OPH, 346). That’s circa 180 A.D., which is 150 years after the religion began. There is not a single reference to this in the entire New Testament itself (for real: see OHJ, 371–75).

Nor does this idea appear in any of our pre-war sources either. No such kin exist—in fact they are conspicuously missing—in 1 Clement’s discussion of church authorities to heed (or even recent martyrs, mentioning Peter and Paul but not James), as I explain in OHJ (308–15). Most peculiarly of all, in the early second century, when Ignatius lists all the historical facts Christians had to affirm as confirming Jesus was a real historical person, his having any brothers never comes up (much less named James, even less as having any notoriety or authority—not even to dismiss or denounce him). We also never hear about this from any references from Papias (OHJ, 323–26), who appears not to have heard of such a thing. When Eusebius writes about the brothers of Jesus, he never has any sources to cite at all, earlier than Hegesippus, despite having used Papias whenever he could. Even the so-called “Infancy Gospel of James” (the Protevangelion) never identifies its named author as the brother of Jesus—that was always a later assumption. And it’s unlikely Josephus ever mentioned this either. No brothers appear in the Major Testimonium; and the Minor Testimonium isn’t likely to have referred to Christians—and even if it did, we cannot establish Josephus knew whether the James he mentions there was a biological or cultic brother of Jesus (OPH, 363–70), just as with Paul. Again, all a result that surprised me.

So the evidence as a whole indicates no Jesus dynasty existed. It was invented in very late fantasy—a century and a half later. It’s not in the NT. It’s not in any first century sources, even those that should mention it. It’s not in any early to mid second century sources either, even (again) those that should mention it. An argument from silence is strong when we have (or also likely should have) documents that should mention a thing, and that expectation is dashed. And that’s where we are. No arguments about, or even awareness of, these “brothers” appear anywhere in any sources for 150 years, despite abundant sources in that period that should have mentioned it. The probability of there even being this dynasty for Christians to argue over is therefore near zero. Which therefore cannot be regarded as “unexpected” and thus improbable on mythicism. This is, to the contrary, more probable on mythicism (OPH, 345–47, 350–51, 373). But a fortiori (being as most generous to historicity as we can), it is no evidence either way. There is thus, again, no inverted likelihood ratio. The probability Jesus existed is simply not affected by these late dynastic legends, precisely because they are so late, and don’t exist at all in our first hundred years of sources.

Conclusion

I had already thought of all these arguments and tested them in my original study, and found them all a wash: they end up carrying no weight, because when we put in all the evidence (direct and background), none of these arguments appeal to any actual evidence that is any more probable on historicity than myth.

• The argument from “dynasties are plausible” collapses from its minor premise being false: there almost certainty was no such dynasty. So there being one only in much later legends can have no effect on the probability Jesus existed. Such late-appearing legends are just as likely on myth as history. “Brother James” is as likely conjured out of fan fiction as any real historical dispute. And there is nowhere further to go with that argument.
• The argument from “worshiping an unjustly executed criminal is implausible” collapses from its major premise being false: peculiarly, Jesus was invented by precisely a people who actually expected and had use for exactly that kind of myth. Even when selling this to Gentiles, the widespread Gentile reverence for the Ancient Oracles of the Jews ensured sticking with and even leaning into this feature, as one could sell it on the authority of those very Oracles, and its internal logic. So this feature is actually entirely expected on mythicism, not improbable.
• The argument from “too soon to have been invented” (or even “no other examples of rapid invention”) collapses on both the minor and major premises being false: we do have such examples, both modern and ancient (all the invented people and public events in the Gospels and Acts, up to and including the Risen Jesus—indeed even a month-long Undead Dinnerparty Jesus); and we don’t have any reason to expect there to be, because what decided when a hero is placed in history is in every known case never decided by how much time has passed, but other factors wholly unrelated. So when the event to be explained (the origin of the cult) is recent enough to match analogous proofs of concept, such a myth is not improbable.

And this is the reason I ended up realizing it’s more likely than not that Jesus didn’t even exist at all: even sensible-sounding, good-instinct arguments for a historical Jesus fall apart as soon as you start digging into their merits. The evidence itself, or the background evidence for antiquity or early Christianity, always unexpectedly makes these things likely on myth, not unlikely. Our intuitions are found faulty as soon as we check the evidence properly, and correctly think about its probability. Indeed, without fail, Everything Falls Apart When You Check. And that is what is improbable for a real historical person or event.