Conversations on naturalistic faith, and the “die for a lie” apologetic

I want do a better job of keeping the cobwebs off this blog, and I’ve decided that I don’t always need to put together elaborate, heavily researched posts to do that. As with the previous post, one simple way to keep things active is to continue sharing some of the interesting interactions that I have in response to the content that others are producing. Toward that end, here are two more recent encounters:

#1 – Naturalistic Faith

If you’re not already familiar with Randal Rauser, I recommend getting to know him. He’s a prolific writer and a thoughtful apologist who employs careful reasoning and regularly campaigns for charity toward those who believe differently – and he isn’t afraid to challenge his evangelical peers. A few weeks ago he tweeted that

Few things are as ironic as a naturalist chiding a Christian about having “faith”

and then followed up on the blog with a little further clarification. At first glance this sounds similar the Turek & Geisler “I don’t have enough faith to be an atheist” line, and I would normally expect Rauser to avoid that kind of rhetoric. Furthermore, as one who sees that naturalism requires little or no “faith” (especially relative to Christianity), I decided to engage. You can read the conversation for yourself in the comment thread of the post, but here’s my TLDR version:

  • Me: Is it still ironic if naturalism is defined in terms of what we are justified to believe exists, rather than what we believe actually exists?
  • RR: Yes
  • Me: I assumed the irony was because the naturalist was asserting the non-existence of non-physical things. Are you saying that the faith is in the superiority of empiricism? I think that there is an inductive argument to be made for the reliability of empirically based ontological claims versus non-empirically based ontological claims.
  • RR: I’ll still ask what justifies that belief. You need to spell out the argument.
  • Me: OK.
    P1: the credence assigned to an ontological claim is rational to the extent that the credence is in proportion to the reliability of the claim.
    P2: ontological claims with higher degrees of empirical reliance have consistently demonstrated proportionally greater reliability, in both degree and frequency, throughout the course of history.
    P3: empirical observations are describable by natural science, so that the degree of empirical reliance for an ontological claim is proportional to the extent to which it can be described by natural sciences.
    Conclusion: it is only rational to assign credence to ontological claims in proportion to the extent to which it can be described by natural sciences.
    RR: I assume you think realism (as opposed to idealism) is justified? In what sense is it justified by way of reliability?
  • Me: Realism requires the least amount of information, so it’s more probable in Bayesian terms, and Bayesian probabilities are ultimately reliable by way of empirical confirmation.
  • RR & JT: I disagree. Idealism is simpler. There are fewer entities.
  • Me: We need to consider the total amount of information, not the number of different entity types. The perceptual information includes realism (the ontological status of our self and other entities). It takes additional information to posit that realism is an illusion, and this additional information is not included in our perception (not empirical).
  • gq: It seems we all agree that realism is “justified as being a properly basic belief”.
  • Me: My acceptance of realism over idealism may be properly basic in practice but I think it is actually probabilistic, as argued above.
  • gq: Even Bayesian justifications rely on “properly basic beliefs”. For example, the reliability of memory.
  • Me: As one who leans toward pragmatism and coherentism, I initially accept the reliability of our cognitive faculties at face value and proceed to look at the study of their reliability to recognize the conditions under which they are more and less reliable, and I use the whole body of data to inform the assessment of reliability without getting hung up on the need for a definitive, indubitable foundation. Regardless, even if we define something in there as requiring faith, I don’t see that this is sufficiently analogous to the religious articles of faith to warrant the claim of irony.

This is almost certainly biased toward doing a better job of summarizing my arguments than the arguments of my interlocutors (I’m noticing a lot more green than black), but there’s a lot more detail in the original conversation if you want to go deeper.

#2 – The “die for a lie” apologetic

Another worthwhile read is the structureoftruth blog. The entire blog is the author working through a very deliberate, thorough and accessible exposition of his belief system in a progression from the ground up. Recently he posted an argument that is similar, but not identical, to what is often known as the “die for a lie” apologetic, wherein the veracity of the resurrection is inferred from the disciples’ willingness to undergo persecution. I find that the typical framing of this argument is flawed in the sense that it creates a false dichotomy that greatly exaggerates the relationship between the adversity and the source of the belief. Since that same approach appeared to be in use here, I decided to chime in. As before, you can read the exchange for yourself, but here’s my TLDR version:

  • Me: I would suggest that it is more appropriate to frame any adversity as being tied to their group identity rather than to one particular belief.
  • SoT: Why believe the disciples faced adversity merely because of their group identity rather than because of their proclamation of the resurrection, and what is there to say that their belief in the resurrection did not become a central part of their group identity?
  • Me: I’m not suggesting that the group identity did not include a belief in the resurrection. Rather, I’m suggesting that the “die for lie” apologetic is an exaggeration. Some reasons to think that “group identity” better relates to the adversity than “resurrection witness” include: historical precedent in general, Jesus’ crucifixion as evidence of pre-existing conflict, Paul does not cite the witness as a reason for his pre-conversion activities, Acts does not focus on the witness in the relevant narratives, and Tacitus cites the eucharist as a motivating reason (though it doesn’t apply to the disciples).
  • SoT: The fact that they persisted in that group identity following Jesus’ death is best explained by their conviction in having witnessed the resurrection. It is pretty clear that this belief was a key part of their group identity after Jesus’ death.
  • Me: The point is that we can’t boil their identity, and the corresponding conflict, down to a single belief. We should acknowledge that there are many factors – belief in the resurrection included – which contributed to the continuance of the group identity and the conflict with other groups.
  • SoT: If the resurrection belief did not originate in a first-person experience, why did the disciples claim it did? Group identity as a motivating force can certainly explain some of the adversity they faced and some of their persistence in the face of that adversity, but to me it seems far from sufficient. And we have no evidence for any origin of the belief in Jesus’ resurrection aside from the first person experiences of the disciples.
  • Me: I think it’s fair to suppose that the group’s belief in the resurrection contributed to adversity (in accordance with prior comments), but it is very different to claim that this requires physical, first-person experiences.I don’t think the gospels and 1 Corinthians 15 are incompatible with accounts that did not originate with the disciples, and even if we grant that they originate with the disciples, there is still the possibility of invention, retrospective mistaken identity, confabulation, or hallucination (or a combination of these).
  • SoT: To be more precise, the inference is from the willingness of the disciples to face adversity for proclaiming that they had witnessed the resurrection, not just from the fact that they faced adversity for believing in the resurrection. I am skeptical that something less than a powerful experience could have changed the disciples and caused the early Christian movement to grow the way it did.
  • Me: I agree that there is in the tradition some evidence that the disciples claimed witness to a resurrected Jesus. However, the phrasing of the argument asserts a very specific, exclusive relationship that is too narrow and overlooks all the other factors at play, so that the relationship is exaggerated. And there was no leaving or returning to Judaism. This was Judaism. The introduction of the belief in Jesus’ resurrection was more a revitalization than a change. These followers had already committed their lives to this group identity, and the drive to maintain that identity is a potent force.
  • SoT: I don’t think it is so narrow as you say, because to me it seems the force of the other factors would be significantly reduced in the absence of the resurrection experiences, so that the other factors are not sufficient to explain what happened. In the aftermath of Jesus’ death, why would the disciples continue to think that they had the true Judaism, when their Messiah had just been killed? Their group identity was “Judaism but done right” and doubt gets cast on the “done right” part by their leader’s death.
  • Me: This group was previously organized around a message that included more than Jesus’ messiahship, which is sufficient to serve as the glue that supports the resurgence that comes with the resurrection belief. They all didn’t simply drop their heads, turn around, and go their separate ways. That said, I’m also inclined to believe that the resurrection belief was a relatively early introduction, and that if it had not been introduced, the group would probably have been a minor footnote to history – much like the Mandaeans. The resurrection belief, coupled with the parousia, renewed their eschatology without having to change their chosen one.
  • SoT: I disagree that the group identity could have been as strong if the resurrection belief came from some other source than eyewitness experience (and I remain doubtful about the plausibility of that belief arising from any other source as well). It isn’t just the belief in the resurrection, but their experience of seeing Jesus risen that motivated the disciples.
  • Me: The “die for a lie apologetic” is typically drawing on the improbability of persons facing severe adversity for the sake of something they knew to be false. The version of the argument summarized in the previous comment is claiming that it is improbable for persons to be in a state in which they are willing to experience adversity, unless they were brought in to that state through direct, first-hand experience. I find that history and psychology shows that this is not at all improbable.

There was also a bit of a side thread where I questioned how he viewed Luke-Acts placing all post-resurrection events in Jerusalem while Matthew (and the hint in Mark) place everything in Galilee, and the likelihood that John 21 is an addition. We didn’t pursue that very far, but I think that a more critical review of the text does not support the notion that there was a strong tradition underlying the post-resurrection accounts in the gospels.

As before, I tried to be fair but this is still probably biased toward more accurately summarizing my arguments. Regardless, I’m interested to hear any other thoughts on the ideas shared in these two discussions.

Share

14 thoughts on “Conversations on naturalistic faith, and the “die for a lie” apologetic

  1. Thanks for the mention! And I think you summarized our discussion fairly. 🙂
    Incidentally, Randal Rauser is from the city where I grew up, and I noticed that a pastor from my home church recently did a podcast with him! Small world.

  2. Hi Travis, I don’t think I’m up to the same level of rigour as you or your “opponents” demonstrate in those discussions, but I’ll offer a couple of comments anyway.

    I agree with you that using the word “faith” to critique naturalism isn’t very helpful Many (most?) naturalists use faith in ways that I wouldn’t agree with when critiquing christianity, and I think that makes some of their arguments ineffective. So I think the same confusion applies the other way round, and more so since naturalists don’t use the word faith at all. It is, I think, a cute barb that antagonises more than sheds light.

    But I think there is still an argument to be made to justify naturalism, and to defend the charge that many naturalists employ arguments against theism that they don’t employ in other areas of life.

    For example, your formal argument doesn’t establish that being “rational” is the only way to arrive at truth and hence that your conclusion (even if correct) is anything to be “feared”. Further, your P2 is only true for certain types of claims – e.g. they don’t give us information on ethics, in fact naturalism tends to lead scientists to deny free will, moral realism and hence moral responsibility, and naturalism has been shown (in some experiments at least) to diminish moral values or responsibility. So you argument may be true for bald scientific facts, but I think it is wrong for many of the more important aspects of life.

    Another example is your statement I initially accept the reliability of our cognitive faculties at face value:. You and I have discussed this before, so I don’t expect we will agree now, but I don’t think naturalism can explain our cognitive faculties, so your initial assumption is (in my view) anti-naturalistic. This page sets out my thoughts on this, developed after our last discussion.

    I also agree with you on your initial objection to the “die for a lie” apologetic. It doesn’t prove that the christians saw a resurrected Jesus, and other factors were undoubtedly involved in their believing despite opposition and sometimes persecution. But the argument is surely more likely, using Bayesian arguments.

    Larry Hurtado seems to have established fairly conclusively that the early christian came to believe in Jesus’ resurrection very soon after he died, and many secular/atheist/agnostic historians accept that the disciples had some form of visionary experiences of Jesus, however they be explained. According to Hurtado, this can be seen first in their worship of Jesus alongside God in a way that other Jews could not do, and which at first they couldn’t explain doctrinally. He argues that this was the result of their resurrection belief, whether that belief was well-founded or not.

    So that is the social and spiritual situation that their willingness to withstand isolation and persecution must be placed in and understood. So we may ask the Bayesian question, was all this more likely if the Jesus hypothesis was true, or more likely if the Jesus hypothesis was false? Now that is for everyone to judge for themselves, but I fail to see how it could be answered any other way that that it was more likely if Jesus really did rise. My argument is this – every reason that you could give for why psychologically the disciples could stand up to persecution if Jesus wasn’t raised can be applied to the case that he was indeed raised, plus there is the reason that he actually was raised as well. So while the argument isn’t proof, it surely raises the probability.

    I further think that your argument that the multiplied probabilities of the resurrection is only 46% is quite mistaken. I am willing to accept your numbers for now, and not argue against them as “structure of truth” has done. Where I disagree is your statement “meaning that the full suite of alternative explanations taken together are more probable than the Christian narrative”. But the full suite of alternatives are not all one explanation, so whatever one of the explanations you favour, it will almost certainly be less probable then the 46%. We can easily see this by supposing there are five alternative explanations, A to E, with probabilities that add to 54%, and each one would be somewhere in the range 5-15% (say). If we choose A, then the full suite of explanations alternative to A (which includes the christian alternative X) will be far more likely than A, or B, or …. So the argument you use could be used to dismiss every one of the alternatives in turn. The only logical way forward is to choose the most likely of the individual options, which is (on the figures we are using) X, the christian explanation.

    Sorry to be so lengthy, but they are worthwhile issues to discuss. Thanks.

    • Hi Eric,
      Thanks for the response. There’s a lot there, so I’ll try to be selective and keep my responses relatively concise.

      your formal argument doesn’t establish that being “rational” is the only way to arrive at truth and hence that your conclusion (even if correct) is anything to be “feared”

      In the context of the dialogue, rational is synonymous with “justifiable” – the challenge was to provide justification for the definition of naturalism that was offered. Do we also need to justify justification? That will just keep us spinning in circles.

      your P2 is only true for certain types of claims – e.g. they don’t give us information on ethics

      The application of P2 to ethics is that non-empirical claims regarding moral ontology are less reliable than empirically grounded ontological claims.

      [quoting me] “I initially accept the reliability of our cognitive faculties at face value”. You and I have discussed this before, so I don’t expect we will agree now …

      Yep 🙂

      Larry Hurtado seems to have established fairly conclusively …

      I don’t think that there’s anything in the work I’ve seen from Hurtado that I find fault with, or that presents much of a problem for naturalistic explanations (except, perhaps for more extreme views like mythicism).

      So we may ask the Bayesian question, was all this more likely if the Jesus hypothesis was true, or more likely if the Jesus hypothesis was false?

      Framing it this way just ignores all the other relevant considerations. What if I asked in reference to the infamous Loch Ness photos: “is the photo more likely if Nessie exists, or more likely if she doesn’t exist?”

      My argument is this – every reason that you could give for why psychologically the disciples could stand up to persecution if Jesus wasn’t raised can be applied to the case that he was indeed raised, plus there is the reason that he actually was raised as well.

      To be clear, I was not rejecting that the apparent adversity of the earliest believers can be taken as some level of evidence for the resurrection. I was pointing out that the evidential relationship is far weaker than is inferred by the way that the argument is presented. The extent of that disparity will ultimately come down to subjective judgments.

      I further think that your argument that the multiplied probabilities of the resurrection is only 46% is quite mistaken

      Someone was reading ahead. I responded to Matt over on that post. Feel free to jump in there if you wish. In the meantime, I’ll just note that there is no contradiction between acknowledging that one explanation is the strongest out of all distinct explanations on offer, and also acknowledging that the explanation is probably wrong. So even if one did judge that the Christian narrative was superior to all other explanations, it does not entail that they would believe it to be true as a whole. And then there’s also the demarcation problem – how specific does any one explanation have to be to consider it a distinct option with its own probability? It all depends on how you want to slice it.

      • Hi Travis,

        “I was not rejecting that the apparent adversity of the earliest believers can be taken as some level of evidence for the resurrection. I was pointing out that the evidential relationship is far weaker than is inferred by the way that the argument is presented.”

        I have agreed with you on this, with the exception of the word “far”. So we both agree that the “die for a lie” argument is neither proof nor totally irrelevant. My comments are simply disagreeing with you about the extent to which the argument can be used to support christian belief.

        “What if I asked in reference to the infamous Loch Ness photos: “is the photo more likely if Nessie exists, or more likely if she doesn’t exist?””

        You have dismissed my argument with an analogy, and I think we both know this can only be a reasonable response if the analogy is a good parallel. I suggest it isn’t in this case, for there are many features which are different. I’ll pick just two.

        In the case of early christianity, the credibility of the evidence has been examined meticulously. The gospels have been analysed to death, and the consensus of secular historians is that they are not inerrant documents by any means, but are reasonable historical reports well grounded in eye-witness stories that have been handed down to the writers. Even though they doubt some or most of the details of the resurrection stories, most historians appear to accept the fact of the empty tomb, that the disciples (i.e. a number of people) had visionary experiences of Jesus and that belief in the resurrection was part of christian belief from the very earliest days and wasn’t a later legend.

        But if we put a photo of Nessie to the same rigorous testing, we would likely get considerable doubts about the reliability and clarity of the photo and the photographer. The events would generally only have been reported by one or two people. Etc. The reliability of the evidence we have would be in considerable doubt.

        Likewise the scientific evidence is different. No-one disagrees that a resurrection is a virtually impossible natural event, but christians, following the early disciples, are not talking about a natural event, but a supernatural one. So if there’s a God, then we have to say that a resurrection is possible, and we have no way to judge the likelihood. But Nessie is claimed to be a natural event, and as such is subject to being assessed by science. And scientists have said that it is unlikely that there is an ancient marine dinosaur in Loch Ness for a whole bunch of good reasons.

        So the analogy isn’t apt. We have good reason to disregard the Nessie photo (unfortunately, I must agree!) but not the same good reason to disregard the resurrection story.

        Now I guess you’ll point to the fact that I have used belief in God to justify the argument which to some degree aims to support belief in God, and thus I have been circular. But I don’t agree. If we are doing Bayesian analysis, belief in God is the hypothesis being tested, and that involves asking these two questions: What is the probability of the resurrection reports if God exists, and what is their probability if God doesn’t exist? To the extent that Pr(G) > Pr(~G), our prior estimate of God’s existence is increased. That is how Bayesian analysis is done.

        Now that isn’t a knock-down argument, for there may other pieces of evidence we could analyse and some would reduce the probability of God’s existence (e.g. the problem of evil) and some would increase it (e.g. credible miracle reports). But the point is that we are discussing THIS argument, and I can’t see how we can’t conclude that this argument increases, whether slightly or greatly, the probability that God exists and Jesus was raised.

        “rational is synonymous with “justifiable” – the challenge was to provide justification for the definition of naturalism that was offered. Do we also need to justify justification?”

        If you are saying that justifiable = rational, then you have ruled out all other ways to justify belief. That is a brave call, and surely requires some explanation and justification!? Philosophers can point to other means, and especially when dealing with intangibles.

        “The application of P2 to ethics is that non-empirical claims regarding moral ontology are less reliable than empirically grounded ontological claims.”

        But when judging moral, supernatural or historical questions, we still have to choose the more justifiable over the less justifiable. So all I think you are saying is that we can believe our conclusions on the resurrection with less certainty than we can believe that 1+1=2 or that the earth orbits the sun. OK, I agree, but that doesn’t get us any way towards resolving the questions we are discussing.

        “So even if one did judge that the Christian narrative was superior to all other explanations, it does not entail that they would believe it to be true as a whole.”

        I think this is quite a surprising statement. If we decided to invest some money on the stock exchange, knowing that all such investments have some risk, who would put their money on anything other than the stock we thought most likely of all the stocks to turn a profit? So if you are saying (just using numbers to be illustrative) that “Even if christianity is 40% probable, and all other beliefs are less than 20% probable, it isn’t necessarily rational to choose the 40%” then I think we are no longer discussing the “die for a lie” argument, but a strange approach to epistemology that entails rejecting the most probable hypothesis. But perhaps I have misunderstood you here?

        Thanks.

      • Eric,
        As is not unusual for us, there might have been some misunderstanding here.

        You have dismissed my argument with an analogy …

        To be clear, the analogy was with respect to the methodology of the argument. I read it as attempting to reduce the whole question down to P(data|resurrection) vs P(data|~resurrection). I would agree that P(data|resurrection) > P(data|~resurrection), but the question is P(resurrection|data), which means that P(data|resurrection)/P(data|~resurrection) must be greater than P(~resurrection)/P(resurrection). In other words, it is not enough to observe that the data is more probable if the resurrection occurred – the probability ratio also has to outweigh the prior improbability of a person being resurrected. So the analogy was intended to serve as an example – we can agree that P(photo|Nessie) > P(photo|~Nessie) is not a complete argument because we also have to consider whether it outweighs the prior improbability of a plesiosaurus living in Scotland.

        If you are saying that justifiable = rational, then you have ruled out all other ways to justify belief. That is a brave call, and surely requires some explanation and justification!?

        What I meant is that you can just substitute the word “justifiable” for “rational” in the argument.

        all I think you are saying is that we can believe our [non-empirical] conclusions … with less certainty than [empirically based conclusions]

        Pretty much. But I don’t think that the distinction is insignificant, and is very relevant to the ‘faith’ comparison that Randal was making.

        I think this is quite a surprising statement … a strange approach to epistemology that entails rejecting the most probable hypothesis. But perhaps I have misunderstood you here?

        If an urn contains balls of 96 different colors, where 5% of the balls are red and each other color makes up 1% of the balls, is it surprising and strange to reject the belief that the next ball to be drawn will be red?

      • Hi Travis,

        “To be clear, the analogy was with respect to the methodology of the argument ….”

        So if I understand you correctly here, you are agreeing that the evidence for the resurrection increases the probability that Jesus was resurrected, but not enough to overcome your prior belief that a resurrection is unlikely. So it isn’t so much the argument that is your “difficulty”, but your prior probability.

        And that is a different question.

        “What I meant is that you can just substitute the word “justifiable” for “rational” in the argument.”

        And I am saying that the words have different meanings. So I think you have to justify your claim that only empirical claims are justified, especially since your argument only establishes that empirical claims are easier to justify.

        “If an urn contains balls of 96 different colors, where 5% of the balls are red and each other color makes up 1% of the balls, is it surprising and strange to reject the belief that the next ball to be drawn will be red?”

        Again, I think it is a dodgy example. If the question you are asking is only “Will this be a red ball?” than you are right. But if you are asking “What colour ball?” then you WOULD pick red. So it depends on the question. And I think the question is “What happened?” If we want to know the TRUTH of what happened, then the resurrection is the most probable option. I think we would only be happy to reject the resurrection if we weren’t searching for the truth, but asking some other question. Historians construct scenarios and decide between them what most likely happened. It would be interesting to see what other scenarios you are considering.

  3. Hi Travis, I posted that comment late at night, but thinking about it when I should have been asleep made me think I wanted to change my last comment on probability, but I can’t find the original discussion I quoted from, where you and someone else discussed the probability of the resurrection leading to the 46% estimate. Where was that?

  4. Hi Travis, thanks for that. I don’t know how I found it before and couldn’t find it a second time.

    I recognise, as you say, that this is just an example. So I’m not taking the numbers seriously, just as indicative, and I’m really wanting to address the ideas behind the probabilities.

    I think I was half right in my post, that the validity of the option you select is determined by what question you are wanting to answer, but I can see that you might want to answer a different question than I do. So let’s go back to your original statement, when you did two things:

    You outlined 4 facts that are often used in support of the resurrection, but which of course are not certain. (These were (1) Jesus was crucified, his burial location was (2) known and (3) found to be vacant and (4) he was seen alive.)
    Then you argue that their conjunction is necessary for the christian narrative to be true, making the final probability less than 50%.

    I want to argue with the logic of the second step.

    Again, it is a matter of what question we are answering. Is it “Is it reasonable to believe in the resurrection?” or is it “Is the entire christian narrative historically true?”?

    If the second, then you are correct, we should multiply all the probabilities. And we should probably add more statements, such as “Were there really two angels at the tomb?” More fundamentally, we’d have to ask “Which christian narrative?” While the gospels agree on the most important facts, they appear to diverge on details. I have seen these details shown to be consistent, but not everyone accepts that conclusion. So you would have demonstrated that there are probably some historical inaccuracies in the resurrection accounts, no surprise to anyone who has read NT historians.

    But I don’t really care about that question, and it certainly isn’t the most momentous one. The important question is the first one, can we believe in the resurrection?

    As soon as we ask this, it is clear that only the first proposition is required (that Jesus actually died) for it to be true that he was resurrected. He could have been resurrected and no-one knew it, or they never visited the tomb but they saw him, or they never saw him but they visited the empty tomb. All of those are possible.

    So the question is, what amount of evidence provides reasonable warrant for belief? I would say a very real experience of seeing Jesus alive and eating as claimed in the gospels would be enough for most people, which increases the probability greatly, to 71% on the figures we have used. We agree the number is not important, but the argument is.

    So it all comes down to what evidence we each think is sufficient. On my criteria, the resurrection is definitely the most probable explanation and more than 50%. Even on your 4 criteria it is still the most likely explanation but less than 50%.

    I suggest all of this “proves” very little, but it does make clearer why each of us concludes as we do. I think that is useful.

    • Eric,
      Thanks for the clarification. In the end I think we can safely agree that it “comes down to what evidence we each think is sufficient”.

      With regard to the simpler, two-step version of the argument, you are correct that it has the effect of increasing the cumulative probability if the individual component probabilities remain the same – but I think that the point of adding the burial and empty tomb components to the argument is to weaken the alternative explanations that don’t account for those, which in turn increases the individual component probabilities for the resurrection hypothesis. At least that’s how I understand the argument.

      As for the previous comment:

      So if I understand you correctly here, you are agreeing that the evidence for the resurrection increases the probability that Jesus was resurrected, but not enough to overcome your prior belief that a resurrection is unlikely. So it isn’t so much the argument that is your “difficulty”, but your prior probability.

      In my original engagement with Matt I tried to make it clear that my objection was not with the position that the data can be used as evidence for a resurrection, but rather with the strength of the argument. But I think that the distinction you make between the argument and the prior probability overlooks the fact that those are both components in the equation. They are two sides of a balance, and just because you cross some threshold on one side (i.e., P(data|resurrection) > P(data|~resurrection)) does not mean that the only thing left to consider is the other side (i.e., P(resurrection)). The total argument weighs them against each other, so that further changes on either side have an impact – you could increase the final probability by increasing P(data|resurrection) / P(data|~resurrection), or by increasing P(resurrection), regardless of where they currently stand.

      It would be interesting to see what other scenarios you are considering

      I agree. These recent discussions have led me to realize that I have mostly traded in general ideas and have not taken the time to work through a more detailed accounting of the naturalistic explanations for the resurrection narrative. I hope to change that at some point.

      • Hi Travis, I think that is a good place to leave that discussion, but I just want to raise one statistical issue that I’m finding myself confused about. (I gained a distinction in Statistics when I studied it back in the stone age, but I’ve forgotten most of it now.)

        In classical probability, which you used to outline those 4 factors which all had to come together, you multiply probabilities of events that are all necessary for the final outcome. i.e. each event reduces the overall probability.

        But if we used Bayesian probability, each piece of evidence (empty tomb, visions) would increase the probability of the final outcome (belief in the resurrection).

        Leaving aside the fact that we’re talking about a contentious event like the resurrection, I’m trying to get my head around why these two approaches seem to have opposite results. I’m sure I’m not thinking rightly about something, but I’m not getting clarity just now. Any thoughts?

      • Hey Eric,
        I think I understand the question you’re asking, and it comes down to what I said in the previous comment about adding evidence to weaken alternate branches. To explain that, let’s use an example tree from Wikipedia:

        For the sake of relating the probability tree to our discussion, let’s say R = resurrection (so C = no resurrection), and P = post-mortem appearances.

        As you can see, the probabilities on the right-most nodes are the multiplied probabilities down each branch, where each level reduces the overall probability of reaching any one particular node. I don’t think there’s any confusion about that.

        Now let’s see where Bayes comes in. If we start down the P(R) branch our probability is only 0.1%, but we can raise it by adding in the appearance evidence (P) even though P(R ∩ P) is lower (0.098%). How? Because the P(R) branch has a proportionally greater percentage of cases that include appearances. So if we add the appearance evidence, we have now weakened (or eliminated) the P(R ∩ ~P) and P(C ∩ ~P) branches. Taking the appearances as certain, P(R) is now increased by 19x, from 0.1% to 1.9%. This can be calculated by using Bayes Theorem, where P(R|P) = P(P|R) * P(R) / P(P) = 0.98 * 0.001 / (0.00098 + 0.04995), but it’s even simpler to just see it as the percentage of appearance nodes that come from the resurrection branch: 0.00098 / (0.00098 + 0.04995).

        Does that clarify?

      • Yeah, I think so, thanks for that. I think the key is that the probability tree has multiple end points, and the Bayes calculation includes several of these end points, as you have shown. Thanks.

      • To be clear, that sounds like the calculation is summing probabilities, and it isn’t. The Bayes calculation restricts the scope of the probability space. Consequently, the nodes (or branches) which remain in the probability space consume a greater proportion of the probability space (i.e., is more probable). Conversely, when you just multiply probabilities (i.e., traverse down a single path of the tree) you are leaving the full probability space intact, such that each traversal places you at a node which consumes a smaller proportion of the probability space (i.e., is less probable).

What do you think?

Loading Disqus Comments ...