This past Tuesday I received my copy of the long-awaited On the Historicity of Jesus by Richard Carrier. So begins my multi-post review of the book.
What This Isn’t
Something needs to be understood before we begin reviewing this book: If you are a Christian and are interested in hearing the arguments against your faith, don’t bother with this book. This concerns a very difficult historical issue. The best case scenario for the author is if he were able to show that the Jesus myth theory has a probability somewhat higher than 50%. In other words, this is almost certainly not something that can be proven beyond a reasonable doubt. What can be shown beyond a reasonable doubt is that the Bible is not 100% accurate, as it is loaded with contradictions, false prophecies, immoral teachings, historical propositions that are falsified beyond reasonable doubt by archaeological data and other historical sources, and doctrines about the nature of the world we live in that are overwhelmingly and abundantly proven false by scientific evidence and common sense. If you want to see the case for that, I’d recommend The Christian Delusion and The End of Christianity edited by John W. Loftus, as well as the resources I cite here and here.
If you are an atheist or skeptic of Christianity, do not use the Jesus myth theory as an argument against Christianity in a debate or with a friend. Here’s why: First, pulling out a doubtful historical hypothesis (which is rejected by a strong majority of biblical scholars, including many atheists and agnostics) only makes our team look bad. Second, establishing mythicism in the space of a conversation or debate is too difficult; it requires addressing and explaining the gospels, Paul’s Letters, Josephus, Tacitus, and so much else besides that even if you are fully able to show that all of those sources don’t count in favor of a historical Jesus you will still be buried in more objections than you can hope to answer in the space of an hour or two. Third, it will always be plausible that there was a historical Jesus. What’s so incredible about Christianity being started the same way many other religions are (with a historical founder)? Why bother with an argument like this when you can draw on so many other arguments that prove their conclusion far beyond a reasonable doubt (like this one)? Fourth, as I will discuss later on, no lay person should have too much confidence in this issue because it is quite possible that some evidence for a historical Jesus has been overlooked, and so even if you read Richard’s book you should understand the highly tentative nature of this conclusion (I’ll tell you what I mean by this later on).
What This Is
First, it may help you somewhat if you understand Bayes’ Theorem before you read this book review, because it plays a big role in this discussion. I’ve written about my understanding of the idea here, and I also have a common-sense and less-mathematical take on how to reason correctly here. In short, my purpose is to evaluate Carrier’s thesis against his own Bayesian reasoning. I will play the role of devil’s advocate (or in this case, Jesus’ advocate) by raising up whatever doubts I have against Carrier’s arguments and weighing the claims against the doubts to see which one is stronger.
One thing that cannot be emphasized enough is the pervasive “Possibility Fallacy” that characterizes this debate: people so often argue that “This is possible” and go on to assume that it is probable. In the back of our minds we must be ready to recognize and call out those who say, “Well, it’s possible that…” with “It’s possible, but is it likely?” Until someone shows that their claim is probable (not just possible) their arguments are worthless. To understand why, understand this: it is possible that the earth is flat. Why do we believe that it is not flat? Well, for one thing, we have trustworthy people who have done experiments (all the way back to ancient times) that show that it is, and for another thing we’ve got photographs taken from space. Of course, a flat earther will interject here by saying: “People lie all the time, we can’t just take their word on the results of those experiments. So what if we have an image of a round earth? Lots of images are faked with photoshop!” What has the flat-earther shown? Nothing much. Sure it is possible that NASA faked those photos and that people lied about the results of their experiments, but given that most people tell the truth most of the time (especially when we have no plausible reason to think they’d be motivated to lie) it is not at all probable. Moreover, given that most people tell the truth most of the time, it is astronomically unlikely that all airplane pilots, NASA workers, and a vast number of other people who would know directly the true shape of the earth would all keep a secret for so long.
When it comes to the question of the historical Jesus we will probably never find anything so overwhelmingly supportive of the myth theory or the historical Jesus theory that only the insane would doubt it. That said, people make the same mistake all the time to a smaller degree. Example: It has been pointed out that it’s a little strange that the central figure of Christianity is named Jesus (a name which means ‘Savior’) unless he was an invention. Historicists typically blow this off with “Well, Jesus was a common name back then, so it is possible that he really lived…” It is possible, but is it probable? Is the name more likely if Jesus was a myth or if Jesus was a historical person? The name need not be wildly unlikely or impossible if Jesus was a real person. If the probability even slightly favors the myth theory, then it is evidence for a mythical Jesus. Perhaps not overly strong or airtight evidence, but evidence nonetheless. Incidentally, I’m only using this an example, we will look at this argument in greater detail to see how it holds up a future post.
Think of it this way: if you walk in a room and see a red jelly bean on the floor, and there are two jars of jelly beans on the counter (one with all red jelly beans and the other with half red and half green jelly beans) it is possible that the bean came from either jar. Nonetheless, it is more likely that it came from the all red jar. It’s not beyond reasonable doubt that it came from the other jar, nor is it impossible, nor is it wildly unlikely. It’s just that the balance of probabilities is on the side of the “all red jar” hypothesis.
What we will do in the coming blog posts is consider how many red jelly beans there are for the Jesus myth theory and how many there are for the historical Jesus theory. To cut along story short, there are facts that are somewhat better explained under each theory, so we must ask who has the larger number of supporting facts and also consider how strong the arguments are: one incredibly strong argument could be better than a hundred laughably weak arguments. Bayes’ theorem, of course, can help us with sorting out difficult cases in which evidence points both ways: it allows us to model the relative probabilities of many pieces of evidence at once and we can also model their relative strengths. Hopefully we won’t need to get too mathematical, but I will be doing my own calculation at the end of this series anyway.
Prior Probability
Probability always involves comparison: if I roll a die, I know that it has a one in six chance of landing on six because by comparing the present diceroll with other diceroll outcomes, I know that ‘six’ comes about one time in every six. When it comes to the existence of Jesus, what other cases are we comparing him with?
Carrier argues that Jesus fits the mythic hero archetype and therefore we should compare him to other figures who fit the pattern (here’s a video that explains the hero archetype). Carrier further argues that since most figures who fit the mythic hero archetype are mythical (not historical) it follows that Jesus probably did not exist. As Carrier is well aware, probabilities can change: the probability I would give for a coin landing heads before I flip it is 50% (prior probability), but once I observe that it is heads (consequent probability), the probability changes to effectively 100% (final probability), and the same might be true with regards to Jesus: maybe we start off skeptical, but still conclude Jesus existed if there’s good evidence in favor of his existence. In other words: If Carrier is right about this, that doesn’t mean we will get all carried away and run out in the streets shouting that Jesus was a myth, it just means the prior probability is somewhat in favor of a mythic Jesus, and the final probability may turn out favoring a historical Jesus if the evidence is there.
But is he right? We can start with the mythic hero archetype itself. First, it seems very loose and vague in some ways, such as the criteria that the hero has “an unusual conception” and “meets with a mysterious death.” Not all the criteria are like that (“attempt to kill the hero as an infant” is both specific and unusual; it is certainly not true about most people) but nonetheless it does raise a little suspicion: to what extent is the archetype a legitimate and meaningful pattern and to what extent is it vague fortune-cookie-prediction nonsense? I have seen with own eyes translations of ancient documents that tell of Hercules and Romulus surviving an attempt on their lives as infants. Hercules and Romulus were sons of God, their bodies went missing and they both survived their own deaths in some way, as Jesus is said to have done and the parallels don’t stop there.* In fact, the same conclusion holds true regardless of whether we use Lord Raglan’s schema or not (in fact, scholar L Matthew White makes use of a similar but distinct “Divine Man Pattern” in Scripting Jesus, which is an excellent book in its own right). It remains true that there are multiple significant parallels between Jesus and other sons of god worshipped at the time, figures that nobody thinks existed.
However, one thing we might question is this: how often are Jesus-like figures real historical people and how often are they mythical? It does no good to find a few examples of real people who fit something like the hero pattern. After all, that amounts to saying no more than that “Archetypal heros are possibly historical because I’ve found a few examples.” Possibility is not probability. How often do real people fit the pattern (or something similar)? To answer this we should try to take a random sampling of figures who fit the hero pattern and see if they are historical or mythical. This is difficult to do, and getting a really large sample is not really possible. Nonetheless, it does seem to me that among the fourteen heroes Carrier lists plus a few others like Caesar Augustus (a real person who was a close match to the pattern) and Kim Jong Il and it is still the case that most characters who fit the archetype (or something similar to it) are myths. Strengthening Carrier’s case, we could cite Herodotus’ account of the God Salmoxis:
“I have been told by the Greeks who dwell beside the Hellespont and Pontus that this Salmoxis was a man who was once a slave in Samos… [Salmoxis] entertained and feasted the chief among his countrymen, and taught them that neither he nor his guests nor any of their descendants should ever die, but that they should go to a place where they would live for ever and have all good things. While he was doing as I have said and teaching this doctrine, he was all the while making him an underground chamber. When this was finished, he vanished from the sight of the Thracians, and descended into the underground chamber, where he lived for three years, the Thracians wishing him back and mourning him for dead; then in the fourth year he appeared to the Thracians, and thus they came to believe what Salmoxis had told them. Such is the Greek story about him.
“For myself, I neither disbelieve nor fully believe the tale about Salmoxis and his underground chamber; but I think that he lived many years before Pythagoras; and whether there was a man called Salmoxis, or this be the name the Getae for a god of their country, I have done with him.” (Histories, Book IV, 93-96).
Here we have an ancient writer telling us that when he heard a story about a godman who came back to life and appeared to his followers, he was unsure of whether such a character really existed. That speaks for itself.
More to the point though, I think other reference classes could be used to get the same approximate result. For instance, Stephen Law has written a paper on the existence of Jesus in which he argues for the ‘contamination principle’:
Where testimony/documents weave together a narrative that combines mundane claims with a significant proportion of extraordinary claims, and there is good reason to be sceptical about those extraordinary claims, then there is good reason to be sceptical about the mundane claims, at least until we possess good independent evidence of their truth.
Law gives an example to illustrate this principle: “If [Sarah and Ted] tell me a man called Bert paid them an unexpected visit in their home last night, I have every reason to believe them. But if they tell me that Bert flew around the room by flapping his arms before dying, coming back to life and turning their sofa into a donkey, well then not only I am not justified, solely on the basis of their testimony, that these amazing things happened, I can no longer be at all confident that any such person as Bert exists.”
I think the abstract reasoning that underlies this principle is that if we judge one part of Sarah and Ted’s testimony [Bert exists] against our background knowledge (The only other place in which we have a means of verifying / falsifying their claims [“Bert flew around the room”], we found they were speaking a probable falsehood) it would seem very uncertain if not outright unlikely that they are telling the truth in this case. Of course, you can’t draw too many conclusions with a sample of one but nonetheless the evidence we have, meager as it is, is purely on the side of pathological lying, insanity, or some other type of deep unreliability for Sarah and Ted.
With the gospels we can do a little better: they said many things which can be independently checked against other historical / archaeological data and against our general background knowledge about the world. Though we cannot check everything the gospel authors say, I think we can be satisfied by looking at a large number of details and events reported in the gospels and seeing how many are true and how many are false out of the ones we can even check at all. We could then use this data to generate a prior probability for Jesus’ existence, our reference class simply being “propositions reported in the gospels.” I have no idea what the results would be, but I suspect it would not be overwhelmingly positive for the historicist theory (we know Matthew made up his infamous mass resurrection in Jerusalem, Mark invented the sea of galilee, the gospels contradict one another wildly, and so on and so forth) though this reference class might give you a decent prior probability for Jesus’ existence. I may attempt something like this at some point in the future and see what results I obtain.
All in all, Carrier’s prior probability of 33% for the historicity of Jesus is reasonable but not entirely beyond challenge, and it may be equally reasonable for us to hold to a prior probability much higher if we use a different reference class such as the one I mentioned.
Brief Update
I was thinking about this at work this evening and came to the following conclusion: if we placed Jesus in a different reference class (by categorizing the ‘life of Jesus’ as an ‘historical detail’ in Mark) the fact would still remain that such a life conformed to the mythic hero archetype, which is more probable if the story was a myth than history. I have done a small personal study in which I randomly selected 24 “details” from the gospel of Mark that one can check against evidence outside the gospels. My conclusion: 60% of details are historical, 40% are not. If we take that as our prior probability, we still have to deal with the fact that Jesus’ life conforms to the mythic hero archetype. If Jesus was a myth, it is almost 100% likely that the stories told about him would conform to the archetype, whereas if Jesus was an historical figure it is somewhat less certain (probably a lot less certain) that his life would conform to it (we could, for example, represent the consequent probability here as ’25 percent’ which I imagine is too high, most known historical figures do not fit it nor do people like John the Baptist and the Apostle Paul). If we do that, it would flip things right back over to over 70% likely that Jesus was a myth. Incidentally, after re-reading page 246, it seems Carrier already accounted for this (how did I miss that before?).
I have also begun to think there is probably more to the mythic hero archetype than I have previously given it credit for. As in: though some of the criteria are fairly non-specific and general, but this is also true concerning the symptoms of a disease. As with symptoms, finding clusters of several symptoms (even if the symptom is broad in its description) can add up to a powerful diagnosis.
Last but not least: if anyone is interested in seeing me publish a blog post about the frequency of Mark’s details being correct, leave a comment.
* Reference: Richard C Miller, “Mark’s Empty Tomb and Other Translation Fables in Classical Antiquity,” Journal of Biblical Literature, v. 129, no. 4 (2010).
