I’ve commented a lot lately in my articles on the historicity of Jesus that critics themselves are now demonstrating why historicity is a bankrupt paradigm: they never have a sound or valid argument for it. Instead, they kneejerk oppose it emotionally, doing no work to even understand the peer reviewed studies questioning it (they rarely even bother reading them, and typically don’t even know what they argue), and then botch facts and even logic when they try to insist mythicists are the ones who don’t know what they are doing. That would be true for many amateur mythicists. But it’s not true for professional, peer-reviewed mythicists. Yet conflating the two is also a typical fallacy critics employ—Matt Kovacs now included (who I believe is the same guy as here or here).
That after ten years this is all they have is now strong evidence that there is no sound defense of the historicity of Jesus. The irony is that these atheists are now acting like Christian apologists: eating their own foot with a misinformed, badly argued traditional position, staking out an emotional rather than a rational argument, and leaning on ad hominem rather than actually checking the facts, while arrogantly claiming the people who actually know what they are talking about (the actual experts who did the actual studies, and thus endeavored to not be misinformed or lean on fallacies) are the ones doing all this. They aren’t. (Nor are they claiming Christianity is false because Jesus didn’t exist, which I suspect is the emotional trigger here.)
Introducing You to the Travesty
Kovacs has blocked my access to his Facebook post (I don’t know why, we’ve never had any hostile interaction) but others can see it and have asked me to respond to it. From them I have screen captures of that Facebook post Kovacs published last month (which I stitched together):
The errors here are numerous and astonishing, ranging from basic failures of fact to botched arithmetic. But it is important to explain each failed step, because what the correct step would have been is what all atheists should be engaging, rather than this nonsense. And this goes beyond just this one question. These are basic failures of critical thinking. You should not be approaching any subject this way.
First, Read the Study
Kovacs clearly has never read either of the peer reviewed studies questioning the historicity of Jesus (by Carrier or Lataster). This is failure mode one: trying to argue against something when you don’t even know what the argument is you’re supposed to be responding to. Which leads to failure mode two: relying on fallacies to “win” (by influencing public opinion) rather than genuine arguments (which would entail getting at what is actually true, regardless of whether public opinion is moved by it). Indeed, his whole post is an ad hominem attack on our competence—and not because attacking someone’s incompetence is ad hominem, but because false claims of it are. What makes his entire argument ad hominem is that his every supposed example of this supposed incompetence is false (which truthfully reflects his own incompetence).
Contrary to Kovacs, I am an expert in all this. Not only am I a “statistics” expert in the only required sense here (I took statistics in college, having already passed college level calculus in high school; I have a dozen college units in electronics engineering, a math-heavy field; and I’ve published multiple peer-reviewed studies in mathematics, including studies Kovacs is talking about), I am also a top-trained expert in Greco-Roman history (including doctoral majors in ancient religion and historical methods). So Kovacs is merely trying to win the argument with this false claim of incompetence, rather than engaging the actual argument. This is a fundamental failure of critical thinking.
And the evidence that Kovacs didn’t read the study he is supposed to be answering is abundant: he aims to proxy-shame us by claiming we didn’t consider margins of error (we did—and if he’d read either study, he’d know that); he doesn’t correctly report any numbers from either study (he thinks I concluded P(historicity|D) is 2%; in fact my lower bound was 6%; and, of course, my actual result was the upper bound, which was 33%; while Lataster’s entire range was nearer 50%); he doesn’t list any of the evidence we advance against the historicity of Jesus (which means he doesn’t know what it is); and the evidence he does list is addressed in these studies, yet he is unaware of how it gets analyzed there, and thus doesn’t know what estimates and arguments he is supposed to be responding to.
For example, Kovacs grossly over-estimates the likelihood ratio for Tacitus, evidently not understanding that you can’t get likelihood ratios like that for wholly dependent evidence—something he would know, or at least know he has to somehow argue against, if he had read any actual study questioning historicity. Likewise the Gospels, where he again doesn’t know why we get a very different ratio, and thus doesn’t know what he is even supposed to be arguing against here. Just touching the surface of this, his ratio of 9:1 would entail that our “multiple accounts” of the life and teachings of Romulus, Osiris, or Dionysus, or even Ned Ludd, Betty Crocker, or John Frum would prove they all existed—exposing a flaw in his reasoning that he would at least know he has to address, had he read any actual study.
So, in short, Kovacs is not even trying. He has literally just phoned this in, oblivious to what he is supposed to be arguing against. That is not the behavior of a critical thinker. A critical thinker knows they have to read the original study—not rumors and legends about it, or even a newspaper article about it, much less its headline; nor even just the study’s own abstract. You have to read the study. And when there are multiple, you have to read them all (for example, in this case, Lataster’s study vetted and corroborated mine against expert critics; so, you kind of need to know how that turned out, don’t you? That is, if you want to know what you are talking about—but, don’t you?). It’s just worse that not only did Kovacs fail at this basic rule of critical thinking, he replaced it with a baseless ad hominem attack instead—he attacked us, rather than our studies. No critical thinker would do that, either.
Second, Check Your Facts (Don’t Make Them Up)
Kovacs claims “80-85%” of scholars “agree” with his likelihoods. This is impossible (no one but us has published likelihood estimates for the facts he lists). It’s also invalid. We can be sure Kovacs just made this number up (including it’s fake range of five percentiles), because there is no such thing as a poll of historians on this issue to consult. Moreover, most of the “scholars” on this specific issue are Christian apologists, and thus delusional propagandists on precisely this issue. That’s not a reliable consensus. And then even if you could somehow locate the non-Christian scholars—how many of them have checked into this specific issue so as to have an informed opinion to count? The reason the “argument from authority” is a fallacy is that merely having a PhD in something does not mean you are omniscient in it. If you have a PhD in Early Modern history but have never studied the subject of Ned Ludd, you cannot have an informed opinion on it. PhDs are not magic. So you have to find the PhDs to poll on it who actually have studied it, and enough to have a reliable opinion on it—and again, you can’t trust Luddites themselves on this, so you have to narrow your poll even further, to scholars not fundamentally faith-committed to believing Ned Ludd existed.
But these polls don’t exist either. So what is Kovacs doing here? His numbers are made up (what does he mean by “80%” of scholars, and what does he mean they assert—and why are the other 20% not convinced of whatever that assertion is—and are they right?). He doesn’t even address whose opinions he is trusting; nor is he investigating whether their opinions can be trusted. I would not be shocked to find that “80%” of these same scholars believe Jesus rose from the dead; and then even Kovacs wouldn’t trust his own cited consensus. But even in the sphere of those allowed to question such claims, the “consensus” in biblical studies is still notorious for turning out to be false (the last hundred years have seen countless shifts in it). So it’s not something you can just “cite.” You need to explain why it is correct on this one issue.
Moreover, the entire point of having a citable consensus is that peer reviewed challenges can ensure it stays honest—but if the consensus is maintained only by ignoring all peer-reviewed challenges to it, it is no longer a citable consensus. And indeed, since the “consensus” is being maintained on this subject by exactly that mechanism, it’s time to call foul on it, not lean into it (see On Evaluating Arguments from Consensus and Things Fall Apart Only When You Check). Indeed, Kovacs may be a rank amateur, but atheists should still be better at this. They can claim they are justified in rejecting all religious claims about the Bible while still being an amateur—so they should be as capable at this as that, or else admit they aren’t competent to claim either. But even experts have done nothing but the same thing Kovacs did: not actually reading the studies, not actually responding to what they say, and, typically, flubbing the math (if they address the math at all). This problem runs all the way down the list of pertinent evidence, too (see An Ongoing List of Updates to the Arguments and Evidence in On the Historicity of Jesus). Meanwhile, for Kovacs’ other made-up polling stat, of “99.9” of scholars “not accepting” mythicism, all the same problems apply—but it’s also dubious given that already more than forty scholars do accept it, as plausible or even correct (see my List of Historians Who Take Mythicism Seriously). How many count-worthy experts does he think there are?
Kovacs actually fabricates all his numbers—none does he justify by any empirical facts or frequency argument. By contrast, in our studies, our estimates for these same numbers are backed by empirical data and references to empirical realities. You can’t refute evidence-based results with made-up armchair results. It’s the other way around. Honestly, that’s Critical Thinking 101. So if Kovacs wants to challenge our estimates, he needs to actually bring some facts in, which will validly change what our own evidence-based estimates are. He needs to engage with the argument. Not just “make stuff up.” That he chooses to do the latter rather than the former is another fundamental failure of critical thinking. To pursue this properly you really do need to read the actual studies, but just for an introduction to the problem here see Doing the Math: Historicity of Jesus Edition.
So, for example, Kovacs says that there is evidence in the Gospels that is nine times more likely to be there if Jesus existed than if he didn’t. What justifies that astonishing assertion? Why nine times and not two or one-point-seven times (or even a hundred)? My study found that stories as heavily mythologized as the Gospels tend to be about non-existent people (no more frequently covering real people than 1 in every 3 times), and that there is not a single passage in them that is any more likely as history than as fiction. Not only do I illustrate the fiction side of this point in detail in chapter 10 of On the Historicity of Jesus, but I demonstrate that any remaining claims to unlikely fictions in the Gospels today all fail to derive from sound logic in Proving History (also peer reviewed, by professors of mathematics and biblical studies), where in fact I show that all other studies of the methods being used to argue this—all of them—come to the same conclusion: they don’t work. The field’s own literature is thus unanimous on this point. That “the consensus” simply ignores this fact is yet another reason why it cannot be trusted in this field. So Kovacs has a lot of work to do here. How does he get this 9:1, when careful empirical studies by experts either don’t get that outcome or don’t logically warrant it?
This problem plagues Kovacs’ armchair list of “evidence” from top to bottom. Not only does he deploy a cherry-picking fallacy (ignoring all the evidence against his position, and listing only the evidence he thinks makes his case—a basic methodology of apologetics, not of a critical thinker), but he doesn’t seem to know what his listed facts are. As I already mentioned, that an author would repeat Gospel claims decades after they were circulating is not improbable on either view; we’d have that whether Jesus existed or not. It therefore has a likelihood of 1:1, not 8:2 (his ratio for Tacitus) or 7:3 (his ratio for Josephus). And that’s even before we get to all the evidence that neither author originally wrote this—the relevant material is all Christian fabrication. That’s true. But we don’t even need it to be. The ratio is still 1:1 even if it’s all authentic. As to why, that is thoroughly addressed in the studies Kovacs didn’t read, including many other studies besides ours (the doubts are even more expansive now than when we published: see the relevant sections of my Ongoing List; more will appear in my next book).
Likewise, the only other “facts” Kovacs lists is “archaeology.” But…what archaeology? There is no archaeological evidence for Jesus. So what “evidence” is he claiming is 6:4 more likely to exist if Jesus did? I suspect he is conflating “his myths include some real places and people” with “therefore they support the existence of every character they portray,” which is not any kind of valid historical reasoning. Otherwise we should conclude Mike Hammer existed simply because his myths get the culture and geography of New York right, or that Ned Ludd existed simply because his myths get the culture and geography of England right (for more on this point see How We Know Acts Is a Fake History). That isn’t how historical reasoning works. It’s certainly not how critical thinking works. As documented by Jan Brunvand in The Vanishing Hitchhiker, myth and folklore routinely acquire “realistic” veneer, as “accurate” local details get added or switched out as the story moves from place to place, culture to culture, era to era. Fiction, too, often employs real people as foils for fictional characters or stand-ins for institutions they represent. By contrast, we do have real archaeological evidence for the existence of Hannibal, or even Pontius Pilate or Caiaphas.
By not checking his facts, but making them up instead, Kovacs has illustrated why historicity keeps being believed for bad reasons. All his listed evidence has an empirically probable likelihood ratio of 1:1; because all of it is as likely to exist whether Jesus did or not. If Kovacs wants to challenge this—at all, much less to get the ratios he just made up (out of, I presume, his “feelings”)—he needs to present evidence (direct or background) that establishes one of the things he listed is somehow unlikely—somehow less expected—if Jesus didn’t exist. And he needs to do that without ignoring the evidence we already presented for our finding that it isn’t. (He also then needs to reintroduce and likewise assess all the other evidence he left out, such as from the Epistles, Acts, or other extrabiblical sources like 1 Clement.) Because that’s what a real critical thinker does. What Kovacs did instead is apologetics.
Third, Don’t Hose the Math
The rest of what Kovacs is arguing is really, astonishingly bad math—in a post claiming to “correct” our peer-reviewed math. That’s embarrassing for him. But it’s also a lesson for critical thinkers: you really need to get this right, or else admit you can’t weigh in with an opinion here. It is particularly funny that Kovacs claims an argument from authority—citing an anonymous mathematician as if they endorsed the content here—but then clearly forgot to ask that authority to actually vet the content here. If they had, they’d have corrected him on quite a few things first. Kovacs does imply even his math is “repeating” things given him by this expert, but if that’s true, his “expert” is actually an incompetent amateur, which might explain a lot—Kovacs gullibly trusted some random crank, rather than anyone with actual expertise. But either way, to just rush in with a completely hosed mathematical argument instead illustrates the lack of critical thinking skills here (because numeracy is essential to critical thinking; but also, you should check with a real expert first, if you are stretching outside your competencies).
Kovacs’ innumeracy is well on display when he calls the H|D in P(H|D) “hypothesis multiplied by data.” That’s bonkers. It’s the hypothesis given the data, a statement of conditional probability, not multiplication. He goes on making that same mistake again when he says P(H) means “prior probability multiplied by hypothesis” (nope) and L(D|H) means “likelihood multiplied by data” (nope). I have to conclude he has no idea what he is doing or what any term in the equation he was given stands for.
This is proved eventually by his also reading the equation wrong, which I’ll get to as well, but first…
The most face-palming moment in his post is when Kovacs inexplicably botches even basic arithmetic. Setting aside the calculation he is making (that’s also bonkers), even if it were correct, his math still isn’t. For example, he claims that on the conservative end (by which he means ‘on the side most critical of historicity’), P(Jesus existed|evidence) equals the simple product of his low-end prior and four likelihoods, hence 0.2 * 0.7 * 0.8 * 0.9 * 0.6, which he concludes equals 0.53 or a 53% chance Jesus existed. That’s incorrect. But again, even if the equation he is depicting a run of here were correct, I cannot work out how he ran the math here to get 0.53. The math he is depicting produces a product of 0.06048, or roughly 6%. Not 53%. This tells me Kovacs lacks even basic numeracy. Even a sixth grader should know that when multiplying fractions, the product cannot be larger than the smallest fraction (if it’s not already obvious why, think about it for a bit). His smallest fraction here is 0.2 (or 1/5). So his output could never be 0.53 (a touch over 1/2). That should have kicked a red flag up in his mind that he’s doing something wrong.
Although the Dunning-Kruger effect doesn’t exist, the Dunning-Kruger phenomenon does—some people who lack the skill to know they shouldn’t be attempting to use a skill they don’t have, will arrogantly proceed anyway and claim victory. And we’re looking at an example of that right here. Critical thinkers pay more attention to their competencies. If Kovacs is so bad at math he doesn’t even know this fractions rule, he should never have even attempted to make an argument like this—certainly without having an actual expert check his work.
But he’s not just terrible at basic arithmetic. Kovacs also sucks at math altogether. His depicted run is wildly incorrect not only because it doesn’t match the equation his anonymous “expert” gave him, but also because it doesn’t even factor in the other likelihoods. There are no likelihood ratios in his run—so he named the probabilities of the evidence on mythicism, and then didn’t even include them in his math. This should also have red flagged him—it should have been obvious he was doing something wrong, if he was expected to estimate some numbers, and not even use them. I suspect this error was due to another: his ratios all sum to 1, which suggests he thinks that’s how likelihoods work. It’s not. Only the priors must sum to 1 (because they divide the probability space among all possible explanations of the evidence). The likelihoods rarely do (and never have to). So you can easily have a likelihood ratio like 0.7 to 0.5 (it doesn’t have to be 0.7 to 0.3). And so on.
Any Bayesian would have caught this for him. The likelihood is simply the probability of the evidence if the hypothesis is true, or if the hypothesis is false. The “likelihood ratio” is then the ratio between those two numbers, whatever they independently are. Most facts are 100% expected regardless. For example, the historicity of Jesus had no measurable effect on the probability that the Roman Emperor Titus would die only two years into his reign. And so that has a likelihood of 100% on both “Jesus existed” and “Jesus did not exist,” for a ratio of 100%/100% or simply 1/1 which reduces to 1 (and anything multiplied by 1 stays the same—hence, no effect). This is actually one of the most important features of Bayesian reasoning: the strength of evidence is always weighed by the difference between the probability of some fact on a hypothesis being true, and the probability of that same fact if that hypothesis is false (which entails some other hypothesis is true)—which captures a fundamental fact of all critical thinking. So to get it wrong is quite catastrophic. Kovacs literally doesn’t know what he is doing here.
And then, the equation. I don’t know what his “expert” told him, but the equation Kovacs says they gave him is correct—it just isn’t the equation he then ran, for some inexplicable reason, but in any case, Kovacs clearly didn’t even look at that equation or understand it, because when he claims to run it, it isn’t what he does. I assume Kovacs doesn’t understand the equation he presents—P(H) * L(D|H) / P/D—has division in it (he also misstyped it, and maybe that was a mere typo, but he means {P(H) * L(D|H)} / P(D); also, L really just means another P). You do run the sequence Kovacs did (yet that, as I just noted, he still fucks up); but then you have to divide by P(D). It is obvious Kovacs has no idea what P(D) is or how to derive it. For those who do want to know, P(D) is just a stand-in for the total probability, which is {P(H) * L(D|H)} + {P(¬H) * L(D|¬H)}. Notice this is where the other likelihoods get in, as {P(¬H) * L(D|¬H)}, meaning the prior probability of mythicism (which is the converse of the other prior, so on his “conservative” prior for historicity of 0.2, this prior is 0.8) times the other run of likelihoods (the likelihood of each item of evidence on mythicism).
So his 0.2 * 0.7 * 0.8 * 0.9 * 0.6 was supposed to be {0.2 * 0.7 * 0.8 * 0.9 * 0.6} divided by [{0.2 * 0.7 * 0.8 * 0.9 * 0.6} + {0.8 * 0.3 * 0.2 * 0.1 * 0.4}], given his chosen inputs (which, as I just surveyed, are also bogus, but here I’m just correcting his math). That’s:
{0.2 * 0.7 * 0.8 * 0.9 * 0.6}
{0.2 * 0.7 * 0.8 * 0.9 * 0.6} + {0.8 * 0.3 * 0.2 * 0.1 * 0.4}
Which is:
0.06048
0.06048 + 0.00192
Hence:
0.06048
0.0624
Which gives us a result of 0.9692 (rounded), or simply 97% (rounded). I ran this out quickly, though, so if I gaffed at all, please correct my arithmetic in comments and I’ll emend my numbers. But I guarantee it’s not 53%. The reason the historicity of Jesus comes out so high (so much higher than Kovacs embarrassingly miscalculated) is really down to his very strong 9:1 ratio for the Gospels, which multiplied the chance Jesus existed nine times in favor, which more than overwhelms a prior of just four times against. So once you then add in the other ratios he invented, roughly 2 times, then 4 more times, then 1.5 times in favor (which all accumulate in geometric progression—e.g. nine times multiplied by four times is thirty six times), you end up pretty close to 100%. Ironically, Kovacs ends by accusing formal expert studies of entering fringe data to force a result (on the GIGO principle of bad inputs gets bad outputs)—but he is the one doing that, right here. His numbers are just made up to suit the conclusion he wants. They have no basis in logic or evidence.
I could correct his other (“liberal”) equation, but I won’t waste your time. If you are keen, the above example gives you all you need to run the correct math on his inputs yourself. But really, it’s the inputs you need to question. Correcting his math is necessary (otherwise he is disinforming the public, and causing maybe hundreds of people to confidently go around thinking this is how you run a Bayesian calculation). But the real issue is his wholly unjustified inputs. By contrast, we justify ours. So any real critic needs to address what we actually argued; not make shit up.
Finally, I also think Kovacs would have made fewer errors mathematically if he had been given the Odds Form (P(H|D)/P(¬H|D) = P(H)/P(¬H) * L(D|H)/L(D|¬H)), but maybe not—he couldn’t even get basic arithmetic right or read an equation properly.
Fourth, Don’t Contradict Yourself
One of the main points Kovacs claims to be making is that a real Bayesian equation would account for margins of error. That isn’t formally true, but it’s functionally true. You’ll notice even his Bayesian equation lacks any margin calculations (so in a strict sense, margins aren’t in any standard formula for Bayes’ Theorem). But because we are working in an area of large uncertainty, we know valid inputs need to be margined. So he is right. And his intended sense is that “we didn’t” do this and so our use of Bayes is bogus. But it’s the other way around. As I noted, this is one of the mistakes he wouldn’t have made had he followed rule one and just read the studies he claims to be responding to. Then he’d know, we did this already. I give what he calls a conservative and liberal estimate for the prior probability, thus setting error margins, just like he did. I also set margins for all the likelihoods—which he didn’t. So despite his supposedly correcting us by adding margins, he forgets his own lesson the moment he gets to the likelihoods, forgetting there are supposed to be margins of error there. He’s the one forgetting this. Not us. We did it right, top to bottom.
This gets to another example of how he obviously never read the studies. Kovacs claims that I concluded the probability that Jesus existed was around 2%. That’s false—twice over (that number in no way comes from me; and it ignores my actual margins). I concluded the probability that Jesus existed was around 33%. That’s the upper margin. My lower margin was then around 6%, not 2% (so I don’t know where he got the 2% from; there is no Bayesian mythicist I know of who has published a concluding margin of 2%, much less under peer review—and remember, Kovacs needs to address the real stuff, not the amateur stuff, if he wants to have any useful point here). Lataster, meanwhile, came to margins closer to 50% (and anyone who wants to assess whether his or my results are the stronger will have to address our respective arguments).
But most importantly, error margins don’t work the way Kovacs seems to think. When we say our confidence interval (expressing the boundaries of our margin of error) is 6% to 33%, we mean, we cannot know what the probability is other than that it is somewhere between 0.06 and 0.33. It is as likely to be 0.06 as 0.33 or 0.12 or 0.23 and so on. We don’t know. That’s the point of an error margin: to capture the scale and range of our uncertainty. So I have to admit the probability Jesus existed could be 1 in 3 (33%). I can’t go around saying it’s 0.06 (and I don’t; and never have). Because it’s an error margin. I do not know it’s 0.06 and not 0.33. All I know is that it is unlikely to be above 33% or below 6%. In other words, the only use of the 6% estimate is to say that I do not believe it is likely to be any less probable than that. And the only proper use of the 33% estimate is to say that I do not believe it is likely to be any more probable than that. And because the latter is what most people want to know, that’s what I should state is what I think the odds are that Jesus existed. If I have time to qualify, I’ll mention it’s a top margin. But I don’t usually need to. The fact is, what my study found is that there is at best a 33% chance Jesus existed. And that’s that.
By claiming we need to use error margins (and thus implying we didn’t), Kovacs not only got our studies wrong (we did use error margins), but then immediately contradicted his own supposed lesson by forgetting he needs error margins on his likelihoods as well (an error we did not make). You could say this also characterizes his entire argument, which is that we are not using math correctly, while in fact he is the one not using math correctly, contradicting himself again (albeit more in an own-goal, foot-eating, face-palming way). But all of this has a serious consequence…
Conclusion
Why “the consensus” is so wrong about the historicity of Jesus is that they are all behaving like Kovacs: flunking all the math, and not reading their own peer reviewed literature, thus criticizing studies they have never read, and consequently getting everything wrong—thus demonstrating the unreliability of their consensus. Which is a fact you all, even as amateurs, can independently confirm (see, for example, the procedure I recommend in Galatians 1:19, Ancient Grammar, and How to Evaluate Expert Testimony, where all you have to do is vet the accessible premises and the logic, skills every critical thinker in the world should have—and should get, stat, if they don’t already have them).
So when Kovacs says this is not how “Bayesian works” (sic), he seems to be the one who does not know how it works. He has no error margins on his likelihoods. He thinks likelihoods must sum to 1. He translates the equation into entirely the wrong calculation (forgetting the divisor). He completely fumbles his own arithmetic. And he never justifies any of his inputs (least of all empirically). But more importantly, he is ignoring the analysis he is supposed to be responding to, which is an even more fundamental epistemic failure. That he thinks this a proper approach demonstrates he is not competent to have an opinion in this subject. He could get competent. He is simply choosing not to. And that is why the consensus in biblical studies so often gets things wrong and thus cannot be blindly trusted as he is claiming it should. Biblical historians are making all these same mistakes. And you can verify that yourself: just apply the above-mentioned critical-thinking skills to every expert critic to date, as you’ll find in my List of Responses.
In the end, when Kovacs claims our studies use “fringe” data (yet does not identify any fringe data in any of our studies), but then himself uses fringe data (like his made-up likelihoods and imaginary archaeology), we can see it is him—the fanatical historicist—who is rigging the inputs to get the result he wants; while we (the actual published experts) are doing the opposite: we are carefully constraining our inputs to explicitly discussed data and measurable uncertainties. So, which of is more likely to be right about the historicity of Jesus? The guy who didn’t study anything relevant, got everything wrong, and hosed the math from top to bottom? Or the guys who got their carefully-argued mathematical analyses published under peer review? Because one thing is certain here: Kovacs’ belief in the historicity of Jesus is here demonstrated to be neither competent nor rational. So. I don’t know. Maybe you should compare that to actual peer reviewed studies on the historicity of Jesus?
When audiences see critiques like this, audiences who know the sophistication of my study (because they’ve read it), the critic only becomes a joke. And then they get angry that everyone laughs at them and ignores them. This is the usual cycle of critique on this topic. The only way to break this cycle is for the critics to get their fucking shit together and start acting like the competent and honest critical thinkers they are supposed to be. I’ve done all I can. It’s on them now to do this right. But first they need to answer these questions of themselves: “Why did you fuck this up so badly? And now that you know your belief was based on a fuckup, why are you still so confident you are right? What’s going on with you?” Because you clearly need to deal with your motivations before you will ever be able do this right.
-:-
Appendix: Bayesian History Is Actually a Thing Now
Kovacs also implied Bayesian reasoning is only being applied to the historicity of Jesus, as if it were some sort of trick. So I will add a bibliography here disproving that. Others before me had already done this; and since my book Proving History, many more have started; indeed, some of those here listed even cite my work on it, so my formalization of the method is already impacting the field (one must also not confuse Bayesian statistics with Bayesian reasoning or mathematics: see Bayesian Statistics vs. Bayesian Epistemology).
First, that all sound empirical reasoning is Bayesian is well argued in countless studies now, these three being just a top-ranking sample:
- Michael Titelbaum, Fundamentals of Bayesian Epistemology 1: Introducing Credences (Oxford: Oxford University Press, 2022).
- Mike Oaksford, Nick Chater, and Ulrike Hahn, “Human Reasoning and Argumentation: The Probabilistic Approach” in Reasoning: Studies of Human Inference and Its Foundations (ed. Jonathan Adler and Lance Rips; Cambridge University Press, 2008), pp. 383–413.
- Thinking: Psychological Perspectives on Reasoning, Judgment and Decision Making (ed. David Hardman and Laura Macchi; Hoboken: Wiley, 2004).
Then the fact that historians should be using Bayes’ Theorem (and/or that the good ones already are and just don’t know it) is a thesis that has passed peer review over a dozen times now, but principally in:
- Aviezer Tucker, Our Knowledge of the Past: A Philosophy of Historiography (Cambridge University Press, 2004).
- Tucker also wrote the upcoming Cambridge Elements volume on this point: Historiographic Reasoning (Cambridge University Press, 2024).
- Richard Carrier, Proving History: Bayes’s Theorem and the Quest for the Historical Jesus (Prometheus, 2012).
- Richard Carrier, On the Historicity of Jesus: Why We Might Have Reason for Doubt (Sheffield-Phoenix, 2014).
- Raphael Lataster, Questioning the Historicity of Jesus: Why a Philosophical Analysis Elucidates the Historical Discourse (Brill, 2019).
- Theresa Heilig and Christoph Heilig, “Historical Methodology” in God and the Faithfulness of Paul: A Critical Examination of the Pauline Theology of N.T. Wright (ed. Christoph Heilig, J. Thomas Hewitt, and Michael F. Bird; Fortress, 2014), pp. 115–50.
- Gard Jenset and Barbara McGillivray, Quantitative Historical Linguistics: A Corpus Framework (Oxford University Press, 2017).
- Barbara McGillivray, Jon Wilson, Tobias Blanke, “Towards a Quantitative Research Framework for Historical Disciplines” in Proceedings of the Workshop on Computational Methods in the Humanities 2018 (ed. Michael Piotrowski; Université de Lausanne, 2018), pp. 53–58.
- Myles Lavan, “Epistemic Uncertainty, Subjective Probability, and Ancient History” in The Journal of Interdisciplinary History 50 (2019): pp. 91–111.
- Nathan Nadeau, “The Philosophy of History and New Testament History: A Survey of the Former with Some Implications for the Latter” in Biblical Interpretation 30 (2022): pp. 374–97 (391–95).
And many historical studies actually employ explicit Bayesian reasoning now, including:
- James Albertson, “An Application of Mathematical Probability to Manuscript Discoveries” in Journal of Biblical Literature 78 (1959): pp. 133-41.
- Winsome Munro, “Interpolation in the Epistles: Weighing Probability” in New Testament Studies 36 (1990): pp. 431–43.
- Raphael Lataster, “Bayesian Reasoning: Criticising the ‘Criteria of Authenticity’ and Calling for a Review of Biblical Criticism” in Journal of Alternative Perspectives in the Social Sciences 5 (2012): pp. 271–293.
- Raphael Lataster, “Questioning the Plausibility of Jesus Ahistoricity Theories—A Brief Pseudo-Bayesian Metacritique of the Sources” in Intermountain West Journal of Religious Studies 6 (2015): pp. 63–96.
- Efraim Wallach, “Bayesian Representation of a Prolonged Archaeological Debate” in Synthese 195 (2018): pp. 401–31 (see my discussion in A Test of Bayesian History: Efraim Wallach on Old Testament Studies).
- Christoph Heilig, Hidden Criticism? The Methodology and Plausibility of the Search for a Counter-Imperial Subtext in Paul (Fortress, 2017).
- Zachary Milstead, “Religion and Arguments from Silence” in European Journal for Philosophy of Religion 10 (2018): pp. 155–69.
- Nevin Climenhaga, “Papias’s Prologue and the Probability of Parallels” in Journal of Biblical Literature 139 (2020): pp. 591–96.
- Nathan Nadeau, “The First Pauline Chronologist? Probably Not: A Review Essay of Ryan Schellenberg’s ‘The First Pauline Chronologist’ from a Bayesian Perspective” in Journal of Greco-Roman Christianity and Judaism 16 (2020): pp. 150–82
- Mark Giacobbe, Luke the Chronicler: The Narrative Arc of Samuel-Kings and Chronicles in Luke-Acts (Leiden: Brill, 2023).
And those are just the studies I know about. There are no doubt more.
I would have a similar problem in relation to your approach to radical criticism (= the view that even the 7 epistles attributed to Paul are entirely forgeries and “fabricated in the Marcion’s school”, cf. Livesey). I mean that, assuming the traditional view on Paul, I agree that you have fixed the best paradigm (minimal mythicism). Only, I would like to know what the best mythicist paradigm would be under the scenario assumed by the radical criticism (even if I know that you disagree with radical criticism). Or if it is even possible to talk still about mythicism versus historicity under the radical criticism (afterall, without more sources dating back to first century, a possible option is even the agnosticism). Thanks in advance for any answer.
Radical criticism does not employ any logical method. It is based on a massive epicycle-build that trounces its prior toward zero, and offers no evidence with a favorable likelihood ratio, much less large enough to reverse its tanked prior.
It’s therefore of no use starting with as a premise. It needs a defense that would convince at least a superminority of scholars first. Only then could one build any further thesis on top of it (like how Christianity might have then begun without a real Jesus).
Proper procedure is to build any such thesis right now on only those facts that are either (a) already a consensus or (b) overwhelmingly provable with evidence (and not brobdingnagian speculation). I explain this in the background chapters of OHJ (Chs. 4, 5, and 7).
Nevertheless, I am funded to develop an article with some examples from Detering next year, which will illustrate why radcrit is a methodological failure. But in the meantime, for the general point, see:
How Do We Know the Apostle Paul Wrote His Epistles in the 50s A.D.?
And:
The Historicity of Paul the Apostle
Also relevant may be:
Was the Entire New Testament Forged in the Second Century?
I remain perplexed why anyone needs to engage in this kind of rhetoric (Kovacs). What is anyone really providing to an audience made up largely of laymen by throwing around Bayes formula and dummy data? You aren’t teaching those people anything they can apply in the future. It’s like when a flat earther sees a picture of something “too far away to be seen on a globe” and runs around delusionally proud of their new fact. The moment someone actually asks, “How far apart are the camera and the object, how high up was the camera, have you ever actually tried to calculate it yourself?” Everything falls apart.
And I say that as a layman. I have some low level understanding of historical methods. I have a 101 level in logic. I can do higher level math, but never had more than a class in statistics (and didn’t know about Bayes until well after college). As such, I do stick to the consensus where I’m aware of it. I still would hold to historicity right now, but it’s more 60-40. And that’s because I just don’t think I know enough to feel validated in accepting the Carrier/Lataster theses.
I try to imagine if I was put in either position. If a strong historicist came up to me really trying to push it as wholly, obviously factual, I wouldn’t just throw a book or even copy-pasted math problem at them. I’d respond with what I feel are the hardest items to overcome (the complete lack of historical credibility in the gospels, Jesus’ life and mission conspicuously missing from Paul’s authentic letters, the complete lack of extra-biblical evidence). On the counter side, if faced with someone strongly mythicist, I’d probably talk about the current consensus among skeptical scholars, James the brother of the lord, and how easily historical people can be deified or at least given magical properties by their followers. My general goal would be bringing down any overconfidence on either side, and I would be trying to give any interlocutor something to chew and and perhaps further research themselves.
It just does no good to say, “Look at this thing that I put literally no effort into but seems to confirm what I already believed beforehand.” Give literally any reason that someone reading it could actually integrate into their own understanding of the topic. Kovacs said that apparently mythicists think the likelihood is 2%. Charitably, maybe he did hear that from some layman mythicist who also doesn’t understand the math and never read any of the professional literature on the subject. Does it really move the needle then to do the exact same thing for historicity? What should the rest of us do when historicists start saying that there’s a 53% likelihood that Jesus existing even under the absolute worst probabilities?
A small error, when you say ‘Kovacs claims “80-85%” of scholars “agree” with his likelihoods’.
Kovacs’ claim is that 80-85% agree with the evidence used to get his likelihoods, not the likelihoods themselves.
Note that if you uncharitably assume that is what he meant, then his statement makes even less sense and is even more embarrassing—because mythicists also agree with the existence of that evidence. So he would then not be saying anything relevant.
The only way to get his statement to be meaningful in this context is if he meant that 80% of scholars agree these facts are positive evidence for historicity, which is just a colloquial way of saying “agree these facts entail such favorable likelihood ratios.”
It’s possible Kovacs is so nutheaded he doesn’t even understand that, but then his post is even more incompetent than I found it.
I had a similar thought to Alick and wondered if you were intentionally misrepresenting his atrociously null argument in order to elevate it to a critquable one, or if you were mistakenly attacking an argument that he didn’t make but one he also did not know he wasn’t making.
Either way, my mind chuckled.
Clicking on the link to Kocac’s Facebook post gives me this error message
“This content isn’t available right
now
When this happens, it’s usually
because the owner only shared it
with a small group of people,
changed who can see it or it’s been
deleted.”
So, I don’t think he has blocked you but probably just restricted it to just friends or a specific group of people.
Dr. Carrier said “blocked my access to his Facebook post” not “blocked me” which either means that they are friends on Facebook but Kovacs went out of his way to specifically excluded Dr. Carrier from the friends able to see the post or that Dr. Carrier expects, as courtesy, that since he mentioned him by name, he should have made it a public post or at least included him in the target audience.
Islam is correct. Facebook does not tell us why access is blocked. It can be for many reasons, not all nefarious. Hence I didn’t say the action was nefarious, only that I didn’t know the reason for it. The natural immediate audience reaction is to ask “Does he have a beef with you?” so I dispelled that with a note that I am not aware of any such reason.
Although it must still be pointed out:
Even limited access settings can be morally problematic here—slander is still slander even when whispered to a single person in a closet. I get the impression Kovacs doesn’t understand that his post is legally libelous and thus immoral, not just erroneous. But to try and slander someone while hiding it being discovered, by any means (even well meaning, such as only communicating that slander to “your friends”), is still censurable behavior.
I don’t know if he will learn this lesson or not.
Dear Richard, once again, that was fucking brilliant. And directly to the point of these people who seemingly want to be recognized by the “stuff” they write but don’t GET IT. I like how like astro science guys like DeGrasse are humble enough to want to be proved wrong so their research can ultimately prove what is correct. Sam Harris introduced me to Ehrman when I was wondering about all this, then through Derek at MV I discovered you and how your approach to historical analysis is, well, scientific in my view.
It’s fascinating and refreshing to see how your arguments through rational empirical evidence blows other scholars out of the water. Wow, “OTHJ just explained something completely different than what I listened to on so and so’s podcast…!” Don’t get me wrong, many scholars bring good learning to the table, but critical thinking is what we need MUCH more of…! Thank you for that.
I have been on this Historical search for Jesus and the origins of the NT for some time now and its amazing how the deeper you go, the more things seem to parallel mythological stories of the Greeks and Romans and beyond. My new thought has been simply WHY??? Why do people in Apologetics, and Theology want to insist that ancient writings in Greek, Hebrew, etc…. are historically true and accurate?
The psychology of religious belief is something that’s wildly fascinating to me. It could be as simple as fear of the unknown and what happens after we die. I like what Yuval Harrari said in Sapiens, the cognitive revolution 50/70K years ago gave us the ability to begin to tell stories. Im glad we can tell stories, just not the kind that terrifies and dominate a culture. Keep up the remarkable scholarship Richard, and thanks for introducing me to your friend Davis Fitzgerald. Y’all keep presenting us “students” with the material and we’ll keep doing our homework….!
Todd from TX
Thank you for the kind remarks.
As to the question of why, the motives are multiple. For the religious, it’s because they have staked their entire hopes and identity on Jesus being real (and not just real, but exactly the sort of person their sect has constructed him to be, regardless of the historical data). For the nonreligious, it tends to be some sort of emotional hostility to “mythicism evangelism,” whereby mythicism is used to mock or refute Christianity or promote wild conspiracy theories that make atheists (and thus, they worry, themselves) look foolish (rather than merely a serious exploration of history), making it again an identity threat issue. It can also be other things (financial motives; social motives; personal motives; academic motives; etc.). I’ve seen various different motivations spill into view with this.
The problem is that delusionality and irrationality are a natural tendency of human beings, so becoming obsessed with angrily defending a false belief is something humans are just prone to. Critical thinking is unnatural; it goes against our innate intuitions, and is literally scary. That’s why so few do it, at all, much less commit to it as a core life value.
“Clinging” is the right word. (Recall Lt. Scheisskopf’s wife in Catch-22, “the God I don’t believe in is a good God, a just God, a merciful God”.) But some who know better also know allowing mythicism risks collapsing their online-course enrollment. Davis and “Simone” seem additionally to wish their investment in Aramaic were not wasted: a Jesus constructed entirely in Greek from the LXX makes Aramaic only a footnote to the origins of Christianity. All such reasons are missed by logical argument.
I haven’t read OHOJ yet. However, I have a guess on why a lot of non-religious people emotionally try to argue against its conclusion. I think a lot of them find the argument from experts’ consensus very effective against fallacious reasoning (which is usually religiously motivated) in a lot of other contexts, most notably in evolutionary biology and multiple topics in cosmology (the earth being an ellipsoid, an ancient universe ..etc.).
I know that you have repeatedly illustrated why this is a huge equivocation fallacy, but I still find this reason very prevalent in most non-religious responses.
Richard, what do you think of the recomputation of the Rank-Raglan priors you use that gives a higher probability of Jesus’ existence? Sources are: https://www.youtube.com/live/gv4bh0qVYgc?si=ASrXWLTSFNb2Gg-j https://hcommons.org/members/kamilgregor/ https://hcommons.org/deposits/item/hc:67611/ https://hcommons.org/deposits/item/hc:67609/
The first is a four hour YouTube video and I could imagine some reluctance to spend that much time on it. The latter two allegedly passed peer review. The speakers say they already had a conversation with you about this, but I am not aware of either a rebuttal of it from you or proposed revised probabilities.
To be clear, if we change the upper estimate of the probability of Jesus existing from 33% to 80% (arbitrary example), we still have reason to doubt it. So I am not saying that the claim in the title of OHJ is jeopardized by this.
I have formal response papers in development.
But the short answer is: they commit errors in set theory and genre theory that nullify their results (in fact their arguments were already refuted in the original study, e.g. in the Alternative Class Objection section of Ch. 7 of OHJ). They are “forcing in” people who don’t belong to the actual set (in effect creating a wholly new set that cannot logically replace the original one).
The only thing with merit is that they make a solid case for adding Alexander, Apollonius, and Mithradates to the actual set. But since I already allowed up to four of its members could be historical, their finding three does not alter my math in any significant way. Insofar as it has any effect at all, it ironically lowers my upper bound to around 25%, but increases my lower bound to around 20%, and so they have only narrowed the error of margins within the wider margin my study already gave. So anyone committed to historicity a fortiori could just stick with my margins; meanwhile, even this result of theirs has no effect on Lataster’s “overgenerous” case for the edge probability being near 50%.
I’m looking forward to your response. I had to look into their characters and they seem to be more of a match than I originally thought. But still very interested in your response. I was reminded of your chapter 7 as I read their findings. I got into with it Hansen after the publication and she seems pretty adamant. But the problem I see is the lack of anything interesting here. Their range swings from unlikely to have existed to very likely and yet for some reason they conclude its likely Jesus existed and stick to their guns that this is a good critique.
Typo: “whom I believe is the same guy”.
The subject of the verb “is” should be “who”, not “whom”. “Whom” would be the object of your belief which doesn’t seem to be the intended meaning here.
Good catch. Fixed.