Review: On the Historicity of Jesus
As a child, a teacher made me agree that even if all the Bible stories were fiction, there must still have been a real Jesus that the legend was built around. This conclusion was unsatisfactory, because how can we say that a man existed, if there is nothing specific we can know about him?
A few years ago, I resolved to find out what we can know about the real Jesus, but finding dependable information turned out to be very difficult. The most recommended author at the time appeared to be Bart Ehrman, but I read his book Jesus, Apocalyptic prophet of the new millennium and was thoroughly underwhelmed. It reminded me not a little of Gavin Menzies’ infamous 1421: the year China discovered the world. The latter sent historians gnashing their teeth, since Menzies’ method is to guess what direction the Chinese fleet may have sailed, conclude the in his opinion most probable path, and then repeat that process over and over again until the fleet has circumnavigated the world and visited every continent. Like Menzies, Ehrman is clearly building a house of cards, where each spurious conclusion is based on an ever growing host of other spurious conclusions, with no actual evidence to ground the ever wilder speculation in fact.
Carrier convincingly mapped the flaws of that approach in his previous book, Proving History, where he argues that historical arguments should be laid out in the mathematical form of Bayes’ theorem. Although I’m not fully convinced that this is always necessary, this method ensures that collective progress can be made because researchers are forced to state clearly what their disagreements are, and prevents them from wasting their time deliberating issues that depend on so many uncertainties that clearly nothing can be said given our present knowledge. In On the Historicity of Jesus, Carrier crucially builds his case around comparing a minimal theory of historicity against a minimal theory of myth, recognizing that there is so little that we truly know about Jesus that we need to start from the rock bottom.
Moreover, unlike even supposedly secular Jesus scholars such as Ehrman, Carrier treats information about Jesus like we would treat information about any other subject. If something impossible is told about Jesus, Carrier concludes that the account is false. Ehrman on the other hand often concludes that we are dealing with a miraculous claim, which lies outside of history and thus cannot be used by historians. This is a bit like a detective who decides to ignore fingerprints since some of her friends don’t like them.
While Carrier has clearly taken care to make his book readable to a layperson, I can’t help but wish he had had a sterner editor. The very large number of parenthetical comments ruin the flow of the book. It is not that Carrier can’t write well, some sections are excellent, but he clearly has an urge to convey every tangential thought and argument that crosses his mind, and he would need someone to force him to kill these darlings for the sake of the greater good.
Bayes’ theorem works by starting with a prior probability that something is true, and then updating that with evidence. From a mathematical point of view this is a little curious, because the prior probability is not mathematically necessary. We could start with nothing, and treat all knowledge and evidence the same. We do not say that the surface area of a cube is the area of the bottom plus the area of all other sides, we simply say it’s the area of all sides. A possible reason for this oddity is practical: Bayes’ theorem has often been used by starting out with a prior probability based on ‘soft’ numbers such as expert opinion, to be updated with more solid (but limited) pieces of evidence.
Thus, Carrier’s use of prior probability seems a little odd. He argues that Jesus is in fact something called a Rank-Raglan hero, and that since all known Rank-Raglan heroes are mythical, we should therefore expect Jesus to be mythical too. My fundamental problem with this approach is that most readers probably do not find this to be the most natural classification for Jesus. That Jesus is mythical is a possible conclusion, but for Carrier it is the starting point. As Carrier points out, it would have been entirely possible to treat all evidence in the same way, and use Jesus’ similarities with Rank-Raglan heroes exactly like any other evidence. I believe that would have been wiser. Since most readers are probably more or less convinced that Jesus was an itinerant preacher—and expert opinion agrees—then that is probably what should be used as prior probability in a Bayesian argument. Carrier argues that it is difficult to estimate the odds that such a character was historical. But this only means that our prior probability must allow for a high likelihood that Jesus was historical—perhaps something like 90%. This could still be overcome by bringing evidence that Jesus was mythical. And that, I believe, is in the end what mythicists need to do. The prevailing view of Jesus is something that needs to be tackled head on.
Unfortunately, I’m not convinced either that Jesus really is a Rank-Raglan hero, or that Rank-Raglan heroes are very unlikely to be historical. I have examined these claims in detail in a separate post. It also seems to me that even if Jesus was a Rank-Raglan hero, that would at most show that the stories about him are fabricated, not that he wasn’t historical. Although one can make the case that legends about historical persons do not take the form of Rank-Raglan heroes, that claim seems spurious on its face. If we had extremely solid data telling us that this must indeed be the case, perhaps we could accept that we do not have a good explanation for how that would be. But we clearly don’t—even in the best case, we only have a handful of examples to compare with.
If Carrier’s use of the Rank-Raglan class in establishing prior probability seems misplaced to me from a rhetorical point of view, I also suspect that treating it differently from other evidence has clouded Carrier’s judgement. Because while he is in my opinion shoehorning Jesus into the Rank-Raglan class in a decidedly apologetic fashion, he is much more sober with his other evidence.
By treating the Gospels without preconceptions, Carrier demonstrates very clearly that these accounts cannot be considered reliable historical accounts, either in whole or in part. These texts are religious fables, invented for the purpose of spiritual education. They were never intended as historical accounts. This conclusion should be obvious, and of course it also explains the countless discrepancies between the Gospels without having to resort to some Shakespearian confusion comedy where authors in the tiny Christian ur-community had access to different sources and were unaware of each other.
This leaves us with Paul’s epistles. And here, Carrier seems to be far too generous to historicity. He allows that the two strained examples historicists have found for their case may carry some weight, which is fine. But he then goes on to concede that Paul’s complete lack of interest in an earthly Jesus might be fully compatible with historicity. In my view, the Bayesian argument from Paul needs to be updated, and it can probably be separated in two parts. First, we can examine the likelihood that Paul would refrain from referring to an historical Jesus simply by accident. We can show that he clearly uses other historical examples, that he clearly considers Jesus to be the most important being and teacher ever to have existed, that he is much concerned with earthly matters in general, and that he frequently would have had occasion to bring up such information in his letters. Thus, I strongly suspect a properly laid out Bayesian argument can conclude that it is practically impossible that such an omission happened by chance. Which leaves us with the second part: considering the possibility that there is some plausible ad hoc explanation for why Paul would purposely avoid discussing the earthly Jesus. Such theories need to be examined one by one. I’m highly doubtful that a defensible explanation exists, and so my personal conclusion is now that there is about a 90% probability that Jesus never existed, and started out as an imagined visionary being. But I must admit that this is not based so much on a careful Bayesian analysis as on the absence of other evidence than Paul, and my subjective judgement of this source. So I remain open to persuasion.
Despite the flaws of this book, I believe it should be a starting point for serious inquiry into the question of Jesus’ historicity. Even if it does not get everything right, at least the framework of the debate is sound. I have little hope that many historicists will take the book very seriously, mostly due to the lack of scholars willing to actually take a secular approach to Jesus and not just state it. But I hope that mythicists and amateur historians from the rationalist community will engage it, since the method in my opinion makes a very good opportunity for a collaborative approach. And that, in the end, is what good science is all about.
Acknowledgement: I have a minor stake in this book; I donated a little to the project, and received an advance copy for reviewing. Unfortunately, I did not have the time to be of much use as a reviewer.