What's the Probability of Resurrection?

Once a week I tune in to the UK Christian radio podcast Unbelievable to hear debates and discussions between believers and non-believers (sometimes between two or more believers). Generally speaking, it has quite a fair format and covers current trends in religion and atheism, from the release of popular new books to the publication of scholarly materials to subjects brought up in the news. The latest episode features Calum Miller and Chris Hallquist in dialogue on an issue that's particularly interesting to me: the probability of the resurrection.

Christian philosopher and apologist Richard Swinburne has attached a 97% probability to the resurrection. How did he arrive at such a high number? Though I have not read Swinburne's book where he makes this claim, the guests and host on Unbelievable dive into some discussion of it. Swinburne uses a mathematical formula from probability theory known as Bayes' theorem. The theorem is a way of calculating the likelihood of something given certain prior conditions. To use an example from a Scientific American article, suppose that 99% of sick people who take a medical test will test positive, and 99% of healthy people who take it will test negative. The doctor has informed you that only 1% of people in the country are sick. You take the test and receive a positive result, and you want to know what the chances are that you are actually sick. Bayes' theorem will give you the answer.

For the sake of space, I won't work out the problem here, but the important thing to understand is that Bayes' theorem depends on knowing certain prior conditions or probabilities. It's simple to ascertain and easy to verify the chances of getting a specific test result, finding out whether someone is sick or healthy, and determining how much of the population is afflicted with an illness. In many cases, these sorts of statistics and data are kept on record by physicians and health organizations, for example. If one miscalculates the odds of being sick after testing positive, it's not that difficult to weed out the mistake from the information provided. But what about when we're trying to determine the probability of something that isn't well known or documented, like a miracle?

One of the most frequent criticisms leveled at Swinburne is that he plays fast and loose with the prior probability of the resurrection. If there is a god, he argues, it's reasonable to assume that it would become incarnate to pass on its teachings to humanity. To validate the incarnation's authority, he continues, god would use a "super-miracle". [link] Someone like Jesus, Swinburne believes, is a prime candidate for god's stamp of approval in the form of resurrection.

On the podcast, Hallquist raises an excellent point that Miller never does address, to my recollection. The mere existence of a god should not lend much credibility to any miracle story, because in addition to competing miracle claims - like the miracles of Mormonism, Islam, Hinduism, etc. - which many Christians reject, there are Deists and many other stripes of believers who do not think their god works in the world to perform miracles. Even if a god exists, why think it acts in nature? Why single out the resurrection as being particularly likely out of all the miracles it could conceivably perform? With all the unnecessary suffering and evil in the world, perhaps the god that exists is an evil god and would have no motivation to raise Jesus from the dead. Why not take the theology of Muslims and Jews into account, who consider doctrines like the trinity and the incarnation to be blasphemous? There's a good argument to be made that, in many ways, Jesus does not resemble the Jewish messiah described in the Tanakh. If the Hebrew god exists, perhaps it sees the Jesus of the gospels as a false prophet rather than its son, and would be against resurrecting him. 

Even if we accept Swinburne's prior condition of  a god existing - for the sake of argument - there are many problems that seem to stop his argument dead in its tracks before ever getting to the resurrection. Remember, all of this must be taken into account to constitute a fair treatment of the prior probability of resurrection. Along with assessing what might count in favor of resurrection, we have to consider all that would count against it, and this is where things become mired in speculation, in my opinion. How do you derive statistical probabilities from things like religious doctrines and facts of the universe that have varying interpretations? If the prior probability of resurrection rests on theistic assumptions, then the entire argument seems to be an exercise in self-justification for Christians, and will be persuasive to no one else.

Like many believers, Calum Miller says he finds naturalistic explanations of the resurrection to be more incredible and outlandish than the idea that god raised Jesus from the dead. Yet some of these are attested in the historical record, like the accusation against the disciples of stealing the body of Jesus. It may be widely dismissed by biblical scholars today because of the guards at the tomb mentioned in one measly gospel, or because the theory doesn't explain the postmortem appearances (whoever decided there had to be one neat, over-arching thematic explanation for everything?), but there are objections to the resurrection hypothesis, too, of a far more devastating nature, in my view.

Several times throughout the podcast, Brierly and Miller bring up what apologists often call a 'naturalistic bias'. They ask Hallquist if he is ruling out miracles altogether from the git-go, on some atheistic commitment to the impossibility of divine intervention. What amuses me about this is how much the Christian explanation is emptied of what little thrust it might have when the assumption of a god is removed. Miller claims that, in his mind, the case for resurrection is strong enough to stand apart from assuming the existence of god, but I have to ask what in particular would compel one to conclude that Jesus was raised from the dead.

The postmortem appearances? Despite Paul mentioning 500 witnesses, we have testimony from exactly none of them, nor do we have their names or any way of verifying their integrity and their story. The gospels mention Jesus appearing to all of his disciples at various times after his death (John 20:26-28, Luke 24:13-16, Matthew 28:16-17), but again we have no written testimony from any of these figures. The only individual to personally report witnessing Jesus after his death is the apostle Paul, though his experience is depicted as a vision (Acts 9:1-9), not quite what the gospels depict for figures like Peter, John, and Mary. Hallucinations are not uncommon among human beings, and with the Book of Acts portraying Peter and Paul as being visited by Jesus upon falling into a trance (10:9-16, 22:17-21), there are alternate explanations to the extraordinary one of actually seeing a man risen from the dead.

The empty tomb narrative? According to the Two-Source Hypothesis, there may actually be just one empty tomb story, in the Gospel of Mark, which was used as source material by the authors of Matthew and Luke. However, the last few verses of Mark are widely recognized by biblical scholars as a late interpolation. [link] The original text of the gospel seems to end at 16:8, when the women flee from the tomb after being told that Jesus has risen, and they "said nothing to anyone." Mark's gospel has been dated to ~70 CE, approximately 37-40 years after the alleged death of Jesus. An empty tomb would also beg of many natural explanations, generally, such as relocation of the body, theft, or even a misidentification of the tomb - all arguably more plausible than resurrection, since history has no short supply of such experiences.

What about the disciples' willingness to die for their faith? Hallquist rightly remarks that the stories of martyrdom are on even shakier ground than the gospel accounts, and brings up Candida Moss' book The Myth of Persecution. The biggest counter-point to this argument all too often goes unstated, though. In short, there is no reason for the skeptic to presume that the disciples knew what they were dying for was false. Many devout religious believers have gone to the gallows for their faith before, including a lot of non-Christians. It could have been, for example, that Jesus' body was relocated or stolen by someone who didn't die a martyr's death, and perhaps wasn't even a disciple. The early martyrs (assuming they were in fact martyred at all) could have remained blissfully unaware of the trickery, going to their deaths no less sincere.

The probability of the resurrection, a priori or a posteriori, is quite low when one seriously considers all the objections to be made. But that's a generous way of putting it. I'm not about to be so bold as to attach any specific number to how unlikely it is, though, partly because I wouldn't know where to begin. There are certainly ways the evidence could be more convincing, like more independent corroboration, dates closer to the purported event, personal written/reported testimonies, a surviving empty tomb with a long history of veneration, and so on. Yet this isn't what we have, nor would it do much to offset the force of the naturalistic explanations, which still far outweigh resurrection in terms of real world experience.