Over the next two podcast episodes, Dr. Calum Miller and I discuss the Prior Probability of the Resurrection. “What the heck is that?” you might be asking yourself. Skeptical arguments against the resurrection of Jesus, for example, the objection that dead people usually stay dead or that miracles never happen, are really arguments that Jesus’ resurrection is supremely unlikely before we look at any of the evidence. Episodes 18 and 19 take these objections to task. Is Jesus’ Resurrection all that unlikely before we look at any evidence?
Dr. Calum Miller
Dr. Calum Miller is a medical doctor, philosopher, and ethicist, currently based in Oxford.
He graduated from the University of Oxford in 2015 with degrees in Medical Sciences (Neuroscience) and Medicine & Surgery. He then worked for 2 years as a doctor in Tameside Hospital, Manchester, while pursuing a part-time degree in Biblical Studies at the University of Manchester. He currently teaches Ethics and Philosophy of Religion at the University of Oxford.
He’s published articles in top journals such as Bioethics, Religious Studies, and Philosophia. He has also taken part in debates, given over 20 conference talks internationally, and written for The Telegraph, The Huffington Post, and The Spectator.
Visit his website here.
As mentioned above, the objection that dead people usually stay dead or that miracles never happen, are really arguments that Jesus’ resurrection is supremely unlikely before we look at any of the evidence. What’s often assumed is that there’s not enough evidence to overcome this initial improbability. That assumption itself can be challenged, and in fact we have plans to challenge it, but today we’re asking, hold on a second, is it actually true that, before we look at the evidence, Jesus’ resurrection is super unlikely?
Links mentioned during the show:
- Dr. Miller’s two appearances on Unbelievable
- Dr. Miller’s lecture on probability theory (all of the lectures in this playlist are amazing)
- Dr. Miller’s debates with Greg Cavin: Did Jesus rise from the dead? & Resurrection: Fact or Fiction
People involved in creating this episode:
- Cameron Bertuzzi – Producer, Host
- Sawyer Hudson – Co-producer, Sound Editor
- Dr. Calum Miller – Guest
To listen to this episode, just hit the play button below. It’s that simple! You can always subscribe and listen on iTunes. All of our episodes are pushed there for free.
Great episode! However, I disagree that the reference class problem is a real issue in Jesus’ case. To begin with, I think the problem in general is a bit overblown. To quote the philosopher James Franklin: “Human life, and for that matter animal life, requires continual judgments of risk on the basis of frequencies — the risk of lions behind rocks, of being waylaid on the way to the shops, of rejection of tenure, and so on. To stay alive and in the game, one must evaluate a good proportion of risks well, which is impossible if one cannot distinguish… Read more »
Hi James, Thanks for your comment. To a large extent I agree with you. Obviously I think we make many judgments on the basis of frequency. But I don’t think the probability strictly *is* the observed frequency – the observed frequency usually just provides compelling evidence such that, if the next observed x is no different in relevant respects from the already observed xs, our epistemic probability that x is F should be roughly equal to the frequency of xs that are F in the already observed cases. The issue here is that we do have extra information about Jesus… Read more »
Hi Calum, thanks for the response! Why exactly do you think *claiming* to be God and *claiming* to perform miracles is relevant in a way that warrants using this narrower class? Putting Jesus’ case aside, there have been countless people in this class who died and, as far as we know, stayed dead. It’s an empirical fact that, if God exists, he has not demonstrated any tendency or desire to treat people in this class any differently with respect to resurrecting them. So I don’t see how narrowing the class to *claimed* Gods/healers is any more justified than narrowing the… Read more »
Well, I take it that one is more likely to be raised from the dead by God if one is indeed God and if one has claimed that one will be raised from the dead. And claiming to be God is good evidence that one is God, even if it is not dispositive on its own. That is enough, however, to suggest that claiming to be God is indeed *relevant* evidence that ought affect the probability here.
Hi Calum, Sorry for the delayed response. Originally, you seemed to be saying that there are narrower classes that have higher frequencies of resurrections, or at least narrower classes for which we don’t know the frequency. You proposed the class “people who claimed to be God.” I responded by saying that we know empirically that resurrections are no more frequent in this class. It seems that you didn’t deny this, but instead argued that belonging to this class is *evidence* that someone belongs to the more more relevant class of “people who are actually God.” Could we use the theorem… Read more »
Thanks for the kind words! So I don’t think frequentism strictly works. But I think frequencies can be good evidence for certain probabilities in some cases. The reason narrowing the reference class makes a big difference is that the reference class becomes a lot smaller, meaning that the evidence from frequency is much less strong. Suppose we have 1 billion people in the relevant reference class (F), and none of them are G. Then we can set an upper bound for the next F being a G as 1 in 1 ,000,000,001. So we have a good argument for a… Read more »
I agree with all that, but I’m not sure how it addresses my criticism. The discussion has gone like this: YOU: There is a narrower relevant reference class – people who claimed to be God ME: But resurrections are no more frequent in that class, so why should we care that Jesus claimed to be God? YOU: But claiming to be God evidence that one is God, and someone who is God is more likely to be raised. So your claims seems to be that “C” (claiming to be God) raises the probability of “G” (being God), and P(R|G) isn’t… Read more »
I think you’d need to include C in every conditional there. And I would also include the rest of the evidence of Jesus’ ministry, namely, the evidence of his not being psychotic, or deceitful, his moral teaching, his teaching of atonement, etc. But if so then yes, you could use the theorem of total probability. And then I would hold that P(G|C&E) is not too low and nor is P(R|G&C&E) where E is the other relevant evidence. So P(R&G|C&E) is not too low and hence P(R|C&E) is not too low.
Wonderful! I think we’re generally in agreement about the methodology and logic of how to consider the prior, and our disagreement boils down to the numbers even plug in. But that’s a discussion for a other day, I suppose.
And you’re right about “C.” I left it out of some conditionals out of convenience (a.k.a. Sloppiness
[…] Part 1 | Part 2 […]
I guess I don’t understand the purpose of this. How are we going to get probability values for stuff like the resurrection? I must be missing some context since Miller was talking about Bayes factors being estimated by some other people. Where did those estimates come from, and how reliable are they? I’m sorry but this podcast seemed to really go nowhere.