How to Improve Bart Ehrman’s Argument Against the Resurrection of Jesus

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

by Neil Godfrey

Matthew Ferguson has posted a very thorough article clearly setting out a weakness in Bart Ehrman’s argument with William Lane Craig over the probability of the resurrection of Jesus.

Simply to say, as Ehrman does, that the resurrection is the “least probable” explanation and therefore it can never qualify as a historical explanation really begs the question. Craig grants that it is indeed the least probable explanation a priori but that the evidence is strong enough to lead the disinterested mind to conclude that it does turn out to be the best explanation for the evidence available. As Ferguson points out:

I don’t think that Ehrman presents the strongest case against miracles (including the resurrection) when he defines them, from the get go, as “the most improbable event.” This kind of definition is too question-begging and it opens the door to the stock “naturalist presupposition” apologetic slogan. The reason we are looking at stuff like the texts that discuss Jesus’ resurrection is precisely to see whether such a miracle could ever be probable.

Ferguson’s article clearly demonstrates the application of Bayes’ theorem in assessing historical evidence for certain propositions and he links to another article discussion the way probability reasoning works in historical studies. (I especially like his opening point in that article pointing out that history is not something that “is there” like some natural phenomenon waiting to be discovered but is a way of investigating the past.) The article also links to another relevant discussion addressing apologist arguments against the likelihood that the disciples hallucinated the resurrected Jesus.

The article is Understanding the Spirit vs. the Letter of Probability.

I won’t steal Matthew’s thunder by singling out here where he believes the emphasis belongs in discussions about the evidence for the resurrection. Suffice to say that I agree with his conclusions entirely.




  • 2017-07-06 11:49:07 UTC - 11:49 | Permalink

    Ehrman was going for a one-punch knock-out, and there was such a punch available. Ehrman didn’t land it, but Craig was vulnerable, as claimed, and his flight into algebra fails.

    Define a hypothesis space. It needn’t be exhaustive. Let a simple “grave robbery” scenario complement “Jesus rose.” Regardless of generalities about miracles, and apart from what these debaters had allowed arguendo, many apologists would agree that a priori, rising is less likely than robbery.

    Observe that the conditionals are equal: either robbery or rising, if true, would perfectly explain the missing corpse. P(observed | hypothesis) = 1, the same for both. Thus the posterior credibility order of the hypotheses is the same as their prior order, p(stolen | missing) > p(risen | missing).

    Ehrman’s intuition was correct, he just didn’t formalize it adequately.

    • Neil Godfrey
      2017-07-07 23:04:55 UTC - 23:04 | Permalink

      As Matthew reminds readers in his article, even a far less likely prior can still turn out to be the best explanation if in fact the evidence turns out to point towards it more than the more obvious options.

    • 2017-07-07 23:41:57 UTC - 23:41 | Permalink


      > even a far less likely prior can still turn out to be the best explanation if in fact the evidence turns out to point towards it more than the more obvious options.

      That’s true, but a substantial part of the article focuses on what Ehrman argued in a specific debate, with stipulations on both sides. So, the merits of the argument that Ehrman actually made then and there is fair game.

      In more general circumstances, there is no reason why a miracle, if it occurred repeatedly under controlled conditions, could not rise above any prior disadvantage short of omission from the hypothesis set (or “zero prior”). One-off events will have a harder time, and for events “in the wild” (e.g. what happened in ancient times) a competent analyst will predictably think of some uncontrolled-for alternative. It only takes one, and often enough just about nothing is controlled for.

  • Scot Griffin
    2017-07-07 16:32:23 UTC - 16:32 | Permalink

    Applying Bayesian methodology to events “occurring” in a literary work of fiction gets you no closer to the truth. In fact, doing so take you farther away from the truth that the literary work is fiction.

    • Neil Godfrey
      2017-07-07 23:07:15 UTC - 23:07 | Permalink

      This is Matthew’s point — and one I’ve tried to make myself from time to time. It is a mistake to naively read a narrative on the a priori assumption that it is documenting real events external to that work. It’s just a story — this most elementary truism is forgotten.

  • 2017-07-07 17:13:45 UTC - 17:13 | Permalink

    A comprehensive account of uncertain reasoning should accommodate what song the sirens sang or what name Achilles assumed when he hid among the women. What confounds Craig may well get you closer to the truth.

    • Neil Godfrey
      2017-07-07 23:09:39 UTC - 23:09 | Permalink

      Sorry but I don’t understand your first sentence. Can you clarify? Thx.

    • 2017-07-08 00:08:32 UTC - 00:08 | Permalink


      The poetic line is Browne’s, best known for having been quoted by Poe (_The Murders in Rue Morgue_) and later used by Richard Threlkeld Cox to help explain his vision of the scope of probabilistic representations of belief. (You can’t condition on contradictions, but you can condition on fictions. Put another way, we can reason about uncertainty in and across all possible worlds.)

      Even if we knew for a fact that the “empty tomb” was a fictive invention, we would still be able to use probabilistic methods to reason about propositions premised on its truth. It is very convenient in debate, for example, to concede a major point to the opponent (OK, suppose there were an empty tomb) and ruin their day anyway (it was more likely robbed than walked out of). The inability to model such a strategy probabilistically would be a major shortcoming.

      • Scot Griffin
        2017-07-08 00:26:39 UTC - 00:26 | Permalink

        Did Frodo Baggins actually disappear when he slipped on the One Ring, or was he just a forgettable schmo? Apply Bayesian analysis and discuss.

        • 2017-07-08 01:47:03 UTC - 01:47 | Permalink

          I don’t see the relevance of your proposal to the merits of what Ehrman tried to argue against Craig.

      • Neil Godfrey
        2017-07-08 01:47:46 UTC - 01:47 | Permalink

        Even if we knew for a fact that the “empty tomb” was a fictive invention, we would still be able to use probabilistic methods to reason about propositions premised on its truth.

        I don’t see the point. If we know a story is fiction then there is no point to going any further to test its “historicity”. We have answered the question before we even ask it.

        If someone doesn’t believe its fiction then the discussion needs to be addressed on whether the narrative is indeed fiction or not. That’s a literary discussion. We can’t assume “for sake of argument” that it is not fiction, nor can we assume it is fiction “for the sake of argument”. Both propositions are simply begging the real question.

      • 2017-07-08 11:57:18 UTC - 11:57 | Permalink

        @ Neil

        > the real question.

        One nice thing about a formal debate is that the “real question” is stated explicitly, up front. In the Ehrman-Craig debate, that question was “Is there historical evidence for the resurrection of Jesus?” according to the report at


        At one point in the debate, Craig proposed that reports of an empty tomb were such evidence. Any demonstrative form of argument has a corresponding Bayesian heuristic argument. We may therefore proceed by cases (a valid demonstrative form):

        Case I. Suppose the reports are false. The parties will likely agree that the reports are so weak as evidence as to be reasonably set aside.

        Case II. Suppose the reports are true. While both rising and robbery become more credible than they were without any reason to suspect that a formerly occupied tomb is now empty, their original order persists.

        Note that one of the case specifications is false, although either is logically possible. If it were invalid to condition on contingent falsehoods, then this form of argument would be invalid, and so there would be a demonstrative form of argument for which the corresponding Bayesian heuristic argument is invalid.

        That would pretty much settle whether Bayes has any comprehensive role in historical reasoning. No. It might yet occupy some niche role, but that isn’t what Laplace proposed.

        • Neil Godfrey
          2017-07-09 00:34:54 UTC - 00:34 | Permalink

          Matthew Ferguson’s point, and mine, too, is that we have no reports of an empty tomb to discuss or assess. Ehrman weakens his position by accepting Craig’s false claim that we do have reports of the empty tomb and resurrection appearances. We don’t.

          (Matthew first makes this point explicit when he addresses the common assertion that Paul is evidence that we have reports of 12 and even 500 witnesses.)

          I used to address this more often some time back (Tim, too) — and you can see the links to the various arguments at http://vridar.org/2011/04/18/the-historical-jesus-hypothesis-does-not-even-rise-to-the-level-of-requiring-investigation/

          The writings we have do not even offer us any reason to begin to think of their narratives as serious history or “records”. The only reason scholars do is cultural and religious heritage.

        • 2017-07-09 09:26:47 UTC - 09:26 | Permalink

          OK, you and Ferguson would have conducted the debate altogether differently than Ehrman did. I accept that.

          Regardless, in light of Hume, it was reasonable for Ehrman to address the specific question at the level of generality that he did. True, the particular reports in hand don’t much support the specific article of faith being debated. True, too, that there is no plausibly envisioned closed set of unsystematically gathered reports that supports miraculous causation over all seriously possible natural hypotheses.

          Given the tactical demands of formal debate, Ehrman’s choice was admissible. Alas, he lacked the technical knowledge to convert his sound intuition about the secular robustness of natural hypotheses into a winning round.

          • Neil Godfrey
            2017-07-09 12:16:39 UTC - 12:16 | Permalink

            I agree that Ehrman’s response was certainly permissible. But his acceptance of Craig’s framework was symptomatic of the most fundamental flaw in biblical scholarship’s study of the historical Jesus or Christian origins. Biblical scholars simply don’t do history or approach sources the way that is normative among competent historians in other fields. Ehrman is trapped in the apologetic set of methods. (His book, Did Jesus Exist? demonstrated most vividly that he has never cracked open a book by historians in other fields on how they go about doing their research and analysis and narrative production.)

            Even in several books written for students about to undertake advanced studies in history one reads of the importance of fundamental questions that need to be established about sources before knowing what sorts of questions one can put to them and what sort of information one can expect them to yield.

            All those rules are violated by most New Testament scholars writing about Christian origins. Biblical historical studies really is an idiosyncratic field that serves to maintain the religious establishment.

  • robert
    2017-07-10 16:48:12 UTC - 16:48 | Permalink

    ” Ehrman weakens his position by accepting Craig’s false claim that we do have reports of the empty tomb and resurrection appearances. We don’t.”

    i think the ehrman alive today believes that women story tellers invented the empty tomb narrative.

  • robert
    2017-07-10 16:48:56 UTC - 16:48 | Permalink

    sorry , i meant they invented the story , not the narrative.

    • Neil Godfrey
      2017-07-11 03:41:32 UTC - 03:41 | Permalink

      He probably does. My point about us not having reports of the empty tomb is to say that we only have an anonymous story that says women discovered an empty tomb. We have no idea who wrote that story and can only speculate as to why. We know nothing about what the women themselves might have said about their experience or the tale that was written about them — or if indeed the women existed outside the imagination of the anonymous narrator.

  • Leave a Reply

    Your email address will not be published. Required fields are marked *