Jesus Mythicism: Two Truths and a Lie

Jesus Mythicism: Two Truths and a Lie July 6, 2018

I’m not sure that I’ve ever actually played the game “two truths and a lie.” I think I have. Then again, maybe I know I have and am lying. Either way, I found myself thinking about the game for some reason as I was reminded of the problematic case that Richard Carrier has made for incorporating mathematical probability (and more specifically a Bayesian approach) into historical methods.

I think that the game helps illustrate the problem.

If one followed Carrier’s logic, each bit of evidence of untruth would diminish the evidence for truth, and each bit of evidence that is compatible with the non-historicity of Jesus diminishes the case for his historicity.

But in history as historians practice it, each claim, each piece of evidence, stands or falls on its own merits. The non-historicity of the cherry tree incident in no way dilutes the case for the historicity of George Washington. There is no need to go back over the evidence and do a recalculation of the case for historicity. That is not just because the impact of that non-historical story is infinitesimally small in comparison with other evidence. It is that the case for historicity is based on the evidence which supports it, and is not diminished by the fact that all famous people also have non-historical claims made about them.

That’s where the analogy of the game seemed to me as though it might be helpful.

The truthfulness of the two truths is not diminished by the one lie, nor is the truthfulness of the one lie improved by the two truths. The two truths are true (if you follow the rules and play the game correctly) and the one lie is a lie, and each stands on its own merits to be evaluated on its own terms.

History works in the same way. And for that reason, all we need for the case for the historicity of Jesus to be solid is some evidence that weighs strongly in favor of his historicity. And we have it, and so in theory, mythicism should simply vanish. But as with all forms of denialism, it will persist regardless of the evidence.

Since I brought it up, why don’t we play “Two Truths and a Lie” here on the blog? I’ll start (no Googling!):

  1. I have been interviewed by E! Online
  2. I have cycled from Dublin to Waterford in Ireland
  3. I speak fluent Albanian

Which is the lie?

Of related interest, see John Loftus’ recent blog posts about Bayes’ Theorem in relation to the claims of apologists who utilize it, such as William Lane Craig and Richard Carrier. See also:

Jesus Mythicism 3: “No Contemporary References to Jesus”


A Doubter’s Guide to Jesus by John Dickson:

A debate between Dennis MacDonald and Richard Carrier:

And Richard Carrier continues writing his own Gospel:


"We don't have to speculate. Physicists and astronomers have proven that 13.7 billion years is ..."

Genealogies and the Age of the ..."
"Unrelated to debate on archaeological objects, the probability for Alien visit to Earth is very ..."

Racism, Colonialism, and Ancient Aliens
"I am a firm believer in Cognitive Science and the AI project. I don't believe ..."

Neolithic Robots
"In an off-beat sort of a way, this reminds me of someone's line (Steve Punt, ..."

God as Parent

Browse Our Archives

Follow Us!

TRENDING AT PATHEOS Progressive Christian
What Are Your Thoughts?leave a comment
  • Phil Ledgerwood

    If I’m following you, you’re basically saying that claims about a person that are non-historical do not constitute evidence that person is not historical. What kind of thing might you consider evidence that a person was not historical that would need to be weighed in the balance? Would it need to be a direct admission that someone was fictional?

    I should add, I am in no way, shape, or form trying to make mythicism sound credible. I’m only asking from a standpoint of genuine curiosity about historical scholarship.

    • John MacDonald

      This is kind of interesting: Carrier says anyone who does not accept his use and interpretation of Bayes is a “Doofus.” See

      • Phil Ledgerwood

        I actually appreciate his restraint in that article. Typically, critics are accused of deliberate deception and malice. It was kind of refreshing to see him refer to them as only being stupid.

        I don’t know what the people actually said that he’s refuting, but for the most part, he makes good points in that article, although I flinched a little when he said this:

        But it can also be used with no numbers at all.

        Oooo! Scary! Magic you say!

        No. Seriously. You don’t even need numbers.

        Thomas Bayes’ paper proving Bayes’ Theorem never once used a single cardinal number (only in an appendix that gave some toy examples of its application). No proof of Bayes’ Theorem ever once uses any number.

        Well, yes, mathematical proofs of mathematical theorems are usually not plugging in numbers into countless examples to show how the numbers work out. The products would be products of the theorem. What would you compare them to? The teacher’s answer key?

        You plug in the numbers for a specific instance of using the theorem, which is what happens in that essay.

        That’s not the only thing wrong with how Carrier uses Bayes’ theorem, but it’s an example from that essay that illustrates that somebody might be making authoritative statements about fields they do not fully understand. If you can imagine such a thing.

        • John MacDonald

          I think Bayes application in the field of history is interesting. Just as we should put a greater emphasis on a “quantitative” approach to history, we should similarly put a substantial emphasis on a “qualitative” approach to math. The question of the quantitative is, at the same time, the question of the qualitative, just as the question of the qualitative is, at the same time, the question of the quantitative. For instance, what ground does the mathematical sentence 3X2=6 stand on? For this ground, the young student often needs manipulatives to visually represent the quantity. When the student uses the manipulatives to make groups of 3, for a total of 2 groups, they understand the total number of things is 6. Try teaching elementary school level math without manipulatives to concretize student thinking!

  • John MacDonald

    If Carrier’s use of Bayes is supposed to lay clear the probability that Jesus didn’t exist, why are professional historians who understand the math not compelled by Carrier’s case? Carrier’s case seems just as unconvincing as it would had he never bothered with Bayes at all.

    • Phil Ledgerwood

      Well, once you rule out the historians who disagree with him who are deliberately lying, afraid of defying the Academy, crazy, or stupid, there may not be many people left.

      You’ve actually touched on a point that I’ve mentioned a few times because, if Carrier limited the scope of his claims somewhat, he’d really be on to something – the use of Bayes’ theorem requires you to quantify what you mean when you say something is probable or not.

      That’s the sort of tool and critique that I think could get a lot of traction. I’m not saying BT is the right tool, but it at least brings visibility into the issue of, when you say something is unlikely, what do you mean by that? What’s behind that?

      Unfortunately, he seems to treat the use of BT as somehow making his claims actually more likely or more valid instead of more clearly delineating his biases. If all he said was, “By using BT, you can see very clearly which sorts of claims I privilege and which ones I don’t. My assumptions are laid bare before you, so you can see how this inevitably leads me to these conclusions,” that would have some punch behind it. Now, people need to figure out how they’re going to achieve that same level of transparency about their assumptions, whether it’s BT or some other thing.

      But instead, the argument sort of morphs into, “BT shows Jesus’ historical existence is unlikely,” which doesn’t follow at all. This is one of the more dubious planks of Carrier’s argument – it’s the assumption that any right-thinking person would quantify and weight the priors very similarly to how he has.

      All he’s actually established with BT is, if you think like he does about the evidence, you’ll come to the same conclusions he has. The use of math does not make this objectively any more correct, though.

      Where he has actual value is in demonstrating that we often make decisions about probability without critically thinking about our biases and making those transparent or demonstrating the impact of such biases on our thinking. In fact, you might be surprised where the probabilities take you once you start quantifying these assumptions. BT is a tool that makes this possible to demonstrate. I don’t know why he doesn’t just stick with that.

      • John MacDonald

        Carrier’s reasoning often gets muddled. For instance, regarding the James passage in Galatians, Carrier says:

        “Given what we have from Paul, this is just as likely, if not more likely, than the alternative reading, because we have evidence direct from Paul that he knows of cultic Brothers of the Lord (as in Romans 8:29 he says all Christians are brothers of the Lord), but no evidence he knows of biological brothers of the Lord, a significantly different category of person. So when Paul says “Brothers of the Lord,” he never says which kind he means; and had he known that there were two different kinds of such brothers, the cultic and the biological, he would need to clarify which he meant. That he never clarifies which he meant, means he only knew of one kind. And the only kind of such brother we can clearly establish he knew, was the cultic. And if even that doesn’t move you, he still doesn’t tell you which he meant; so you can’t otherwise claim to know.: see

        – Carrier’s conclusion “That he never clarifies which he meant, means he only knew of one kind” is in no way established from the premises of the argument he has laid out.

        • Phil Ledgerwood

          I think he just needs to give up drawing conclusions.

          • John MacDonald

            Yeah, Carrier has this odd habit of leaping from (i) the idea that his interpretive model is “compatible” with the evidence to (ii) the conclusion that his interpretive model is the “most likely” explanation of the evidence.

  • It is impossible to raise anyone from the dead on Earth but if you believe Elijah raised the dead and that justifies mass-murder of prophets of the god of storms (Ba’al) you have no excuse not to believe Jesus raised the dead.

  • John MacDonald

    Two thoughts:

    (1) I watched the John Dickson video Dr. McGrath provided above. I liked the part about Jesus radicalizing the Jewish concept of love to include the outcast and the sinner. On the other hand, Dickson seems to want to sneak in the plausibility of the historicity of the healing miracles of Jesus (in no small part because of their early and multiple attestation), and poke a little fun at Paula Fredriksen’s skeptical interpretation of the healing miracles (video time index 28:51 ff). Where Dickson’s argument here sort of falls apart here is when he points out Q’s Jesus interprets the healing miracles as previews of the coming Kingdom where there will be justice and wholeness (video time index 29:43; cf Matthew 12:28, Luke 11:20). The general hermeneutic rule of thumb is that if something in the text (in this case the healing miracles) shows theological motivation (similar to Matthew presenting Jesus as the New and Greater Moses), we bracket its historicity (it may or may not have happened) because the author would have had reason to invent it. Since the healing miracles serve the function of previewing the Kingdom in the text, they may or may not go back to the historical Jesus. I’m also a bit hesitant to agree with Dickson that we can be sure that there are as many sources underlying the gospels that he thinks there are. For instance, there may be a separate ‘M’ source underlying Matthew, but it could also be that Matthew simply invented the material unique to him. We know in some instances Matthew was simply inventing stuff, like the “Jesus is the New and Greater Moses” business, so he could have been inventing the other stuff unique to him as well.

    (2) Neil Godfrey at Vridar has posted a response to Dr. McGrath’s above post here:

  • Nick G

    I first came across Bayes’ theorem in the context of medical diagnosis (I’m not a medic but a computer modeller, I was working on a medical expert system). In that context, the question is the probability that person A has condition X, given evidence E – typically a test result, perhaps along with various signs and symptoms, The prior probability will typically be the prevalence of the condition among some population A belongs to. The choice of population, hence of prior probability, is not trivial: suppose A is a woman of 57, living in London, working as a barrister, teetotal, and of Afro-Caribbean descent. Any of these factors (sex, age, location, occupation, alcohol intake, ethnicity) could be relevant – among many others – depending on what condition X is; the medic (or expert system) will need to take into acount what statistics are available – the number of 57 year old teetotal female barristers of Afro-Caribbean descent living in London is almost certainly too small to provide a sample of useful size. The factors could also affect the likelihood of false positives andor negatives in test results. So even in this type of case, Bayes’ theorem isn’t simple to apply; but crucially, there will be a body of relevant statistical results, collected with the express purpose of aiding decisions of the type to be made. Things could scarcely be more different in historical research, even where the question is of the type both Carrier (and Craig) want to apply Bayes to – does story S recount a real event or not – which is a tiny part of what real historians are interested in. The choice of reference classes to provide a prior probability and the weight to be given to each piece of evidence is bound to be arbitrary (like Carrier’s choice of Rank-Raglan heroes – it’s not even as though the validity of this class was a consensus among relevant experts), and the choice of relevant evidence itself is a matter of judgment. In short, the attempt to apply Bayesian reasoning in this context is a prime example of inappropriate quantification.

    Carrier claims that Bayesian reasoning is simply a formalised version of all valid historical reasoning, but this is nonsense. Typically, historians are seeking to understand or explain bodies of evidence with multiple unique or idiosyncratic features, in terms of some set of underlying mechanisms, events or processes that could have produced them. (Sometimes they attempt to formulate general laws of historical development, but that’s not particularly relevant here.) To put it another way, they are generally looking for the best explanation of a body of evidence. Choosing between alternative explanations will typically involve considerations of the plausibility of the events or processes each would require, but only rarely will this involve numerical probabilities – at most, there will be a partial order of likelihood assigned to the hypothesized events or processes. This may well involve considering whether we know of cases where similar events to those hypothesized have occurred – but it makes no sense to ask in how many cases they have not occurred and hence calculate a numerical probability. So what Carrier needs to do to establish mythicism as a serious historical hypothesis is to construct an explanation of how the body of evidence we have could have been produced without a real person’s life being at the base of it, which is at least within range of being as plausible as the historical-Jesus hypothesis, that there was such a person, subsequently heavily mythologised – and we know that such mythologisation does occur.This, in the view of almost all relevant experts – including atheists, agnostics and observant Jews – he has not done.

    • John MacDonald

      Wouldn’t you say, in a sense, historical analysis is quantitative, in that it tries to infer what “probably” happened (with a relative range of 0.51-1, respectively), and to rule out what “probably” didn’t happen (0.49-0)?

      • Nick G

        No, I wouldn’t. Absoutely nothing is gained by sticking numbers on a qualitative judgement when there’s no justification for them. What could such numbers refer to in the case where the question is specifically, “Did [event E] happen?”? Either E did happen or it did not, so there’s no sense in which it happened 51% of the time, or to the extent of 0.51. Bayesian probabilities are sometimes interpreted as the odds you would require to bet on something, but those say more about you than about the event (or non-event) in question. What “E probably happened” means to a historian is something like: “We have to make fewer or less far-fetched additional hypotheses to accept that E happened than to reject it”.

        • John MacDonald

          So you wouldn’t say, when trying to infer whether an historical event happened, that it was”likely, possibly, or unlikely to have happened?”

          • Nick G

            Yes, of course I might. What are you on about? “Likely” means much the same as “probable”. Neither requires or implies quantification. Do you really think people didn’t talk about whether something was “likely” or “probable” before decimal fractions and probability theory were discovered?

          • John MacDonald

            I’m not sure how you understand the distinction between “quantitative” and “qualitative” reasoning? For instance, if I know your grandmother “usually (but not always)” calls you on Sunday, without knowing any other information, I can reasonably “pre-dict” that she will call you next Sunday, and “post-dict” that she called you last Sunday. To me, this mainly seems like “quantitative” reasoning, with something like an approximately 2/3 likelihood inference that your grandmother called/will call on the two respective Sundays. The math is approximate, but certainly you would say people engaged in mathematical reasoning even before the being-made-explicit of certain mathematical models? Maybe you could define what you mean by “quantitative reasoning” and “qualitative reasoning,” and give some examples so I could better understand where you’re coming from?

          • Nick G

            To me, this mainly seems like “quantitative” reasoning, with something like an approximately 2/3 likelihood inference that your grandmother called/will call on the two respective Sundays

            Why? Where does the 2/3 come from? Until I understand why you have this peculiar need to attach numbers to all judgements of probability, likelihood or plausibility, I’m not going to make any further efforts to correct it.

          • John MacDonald

            You do realize that when we teach children probability, we do it in Math class?

          • Nick G

            So what? There is indeed a mathematical theory of probability. That does not imply that anytime the word “probable” or “likely” is used, a number needs to be, or can usefully be, attached to it. Often, as is the case with your “approximately 2/3”, there is no justification whatsoever for doing so – and that is inappropriate quantification. It would be a good idea to teach children when it is and is not useful to use numerical terminology, as part of learning about probability.

  • Nick G

    I’m sure I posted a comment here, which has disappeared. Since I’m clearly not banned (as I can post), and I’m pretty sure there was nothing objectionable in it (it was just generally sceptical about the use of Bayes’ Theorem in historical research), I wonder what happened.

    • Not sure how it ended up in the spam filter, but I’ve added you to the trusted user list, and so it shouldn’t happen again. So sorry!

      • Nick G


  • John MacDonald

    My hypothesis is that the historical Jesus existed. I know most people believe this, but I don’t think the average, every day person is prepared to defend this position against mythicist objections. For instance, how many people are aware of the James passage in Paul, and how much weight it carries?

    I always try to play Devil’s Advocate and come up with counterexample to test my hypothesis.

    For example,

    (1) One possible counterexample to the historical Jesus hypothesis that I came up with is that Paul seems to thinks seeing the risen Christ is a necessary (but not sufficient) condition of being an apostle (1 Cor 9:1). This seems odd on the historical Jesus hypothesis. Would the twelve have stopped being apostles if they never had the revelatory experience described in the Pre Pauline Corinthian Creed?

    (2) Why did Paul not have to go through the lengthy mentoring process Jesus required of his disciples in the Gospels? A possible explanation is the original Christians had no such mentoring process as portrayed in the gospels, and were just a revelatory cult.

    Asking questions like these have helped me better understand my historical Jesus hypothesis and become more confident in my position.

    • These questions might perhaps lead one to question the existence of a “council of the twelve” prior to Easter and the church. But how would it cast doubt on the historicity of Jesus, if at all?

      • John MacDonald

        Initially I thought (1) might suggest that since apostleship seemed to be (partially) dependent on having had a revelatory experience of Jesus (although not completely dependent, since Paul reported the 500 as also having a revelatory experience also without being apostles), this would fit in nicely with the idea that the original Christians knew a celestial Christ simply through revelatory experiences. But, as you say, all 1 Cor 9:1 is suggestive of is possibly the nature of apostleship post Easter, which doesn’t really have anything to do with whether Jesus existed or not.

        As for (2), it is odd to me that Paul became an apostle without having to do any job shadowing like the disciples did in their mentoring process with Jesus, but this “perceived oddness” requires psychologizing what the apostles would or wouldn’t do when accepting new members post Easter.

        Anyway, as I said, playing Devil’s Advocate, these are the two objections to the historical Jesus I came up with today, and I am confident they can be resolved without any recourse to adopting the Christ Myth Theory.

  • John MacDonald

    I don’t see how literary allusion in the New Testament is evidence that the writer simply invented the material out of whole cloth. And we have a specific instance involving literary dependence that nevertheless did actually happen:

    Paul’s conversion story in Acts is modeled on 2 Maccabees 3’s story of Heliodorus and Euripides’ “Bacchae.” Now, if we didn’t have the conversion story in Paul’s letters, mythicist reasoning would seem to entail that we would have reason to believe, purely from the literary dependent account in Acts, it is more likely than not that Paul’s vision never happened. But such reasoning would lead to the wrong conclusion, since Paul actually did have such an experience.

    Literary dependence/modelling is not evidence that the pericope never happened.

    • The Mouse Avenger

      Precisely! 🙂