This is a fantastic book for serious history geeks/nerds/wonks that tackles the eternal question of how we can actually know anything about the past. History is getting slightly easier to pin down in an age where everyone has a camera, but what about ancient history? How can we know anything about what really happened thousands of years ago? The answer is Bayes Theorem. If you haven’t heard of if before, it is the mathematical and logical formulation of all scientific inquiry. Bayes Theorem is an equation that seeks to answer the following question:
“Regarding any unexplained piece of evidence, how likely is it that my preferred hypothesis is true in comparison to all competing hypotheses given everything we know about the world?”
All valid methods of logical deduction reduce to Bayes Theorem. Any time you make a judgement about the likelihood of a claim being true, your brain is using a loosey-goosey, guesswork version of Bayes Theorem. Any disagreement between opposing hypotheses can be resolved through Bayes Theorem. It is the single most powerful theorem in all of mathematics, and widespread application of it would result in a lot less confusion about the nature of truth.
Now if you take issue with any of the assertions I just made in the last paragraph, then Proving History: Bayes’s Theorem and the Quest for the Historical Jesus is the book for you. Dr. Richard Carrier lays out a stunningly detailed, obsessively thorough, and nigh irrefutable case for the use of Bayesian reasoning as the only valid method of historical inquiry.
Escaping from ignorance requires acknowledging that certainty is an illusion. No matter how we feel, we are only ever as certain as our current knowledge allows us to be. If we are using bad information, our conclusions will be bad even if our logic is correct. In reality, our certainty about anything is plotted on scale between 0 and 1, but it is never actually 0 or 1. It is always a degree of probability. Carrier argues, with tremendous success, that Bayes Theorem allows us to cut through the foggy guesswork of emotions and intuition by assigning real numbers to those probabilities.
In other words, Bayes Theorem counteracts Confirmation Bias and shows us with mathematical precision what we can actually know versus what we don’t have enough evidence to support. This allows for real problem solving, progress, and evolution. It focuses the conversation on the areas of real disagreement, while putting to rest arguments over matters where the evidence is genuinely conclusive.
It is very important to note that the topic of the Historical Jesus is entirely incidental to the point of Proving History. Carrier has written a totemic book on that subject called On the Historicity of Jesus: Why We Might Have Reason for Doubt. If you are interested in Jesus studies on any level, regardless of religious affiliation, this is not just a must-read book… It is a book so essential that you cannot knowledgeably speak on the subject without considering the evidence it presents. I’m not saying you in any way have to agree with Carrier’s conclusions, but you cannot credibly ignore his evidence.
All that said, Proving History is intended to lay out a logically sound method of historical inquiry which he then deploys in his other books. Since this book is a deep dive into the method in all of its delicious detail, the target is ultimately irrelevant. The method can be applied to any form of evidence-based inquiry.
This is not casual reading. This book assumes that you have a college-level understanding of probability theory, statistical analysis, and common historical research methods. This was written by a serious historian as a reference tool for other serious historians. Carrier acknowledges that some of the mathematics required are challenging, but he rightly points out in his firm yet affable style that if you’re serious about truth…you have to do the math.
The common objection to Bayes Theorem is that it requires you to provide specific probability values for things that often feel very arbitrary. Since the inputs are somewhat arbitrary, doesn’t that make the outputs arbitrary, too? How can we trust the conclusions if we’re just guessing about some of the inputs?
It is true that your results are only as trustworthy as the evidence that supports your inputs…but this is how your brain already works! This objection is saying, “I don’t like the fact that there is guesswork involved and I would prefer a method that doesn’t require guesswork.” Yet there is no such method! All methods for determining truth in complex systems rely on some degree of guesswork. The only difference is that Bayes Theorem is honestly, plainly showing you where the guesswork is happening while all other so-called methods are hiding, obscuring, or outright lying about it. The only way to minimize guesswork is to shine a light on the exact step in the process where the guess first becomes an input, which is precisely what Bayes Theorem does. Once you know that, then you can refine the certainty of your guess by referencing specific evidence. The more evidence you can compile, the more credible your guess becomes, until it reaches a point where any reasonable person who looked at the evidence would be forced to agree that yours was the best possible guess.
Carrier also digs into a favorite challenge of mine, which he calls the Threshold Problem. This is simply the question of how much evidence do we consider an appropriate baseline for establishing if a historical figure really lived or was just a myth? Why are we so confident Socrates was a real person but King Arthur was merely a legend? The answer lies in examining the contextual evidence of the time and place they are said to have lived.
Socrates lived during a thriving period of Ancient Greek history (470/469 – 399 BCE) where there is a great wealth of surviving evidence in the form of artwork, documents, architecture, etc. Even though Socrates himself did not write anything, he was written about by multiple contemporary sources, including Plato, Xenophon, and Aristophanes. According to Bayes Theorem, our background knowledge of how the world usually works states that this kind of multiple, independent attestation is reasonably expected if Socrates was a real man, but would be somewhat unusual if he was just a myth or fictional character.
In contrast, most of the colorful stories and popular imagery that comprises the modern King Arthur legend can be traced back to the book History of the Kings of Britain by Geoffrey of Monmouth. This was written in 1138 CE, a full 500 years after Arthur supposedly lived! There are no known sources prior to that date to corroborate the stories, and 500 years is a long damn time. Bayesian reasoning dictates that any story so far removed from the supposed events is not usually a credible history, so we are safe to conclude that the figure we know as King Arthur is just a myth unless new archaeological evidence is discovered to refute that.
So while establishing a credibility threshold may be somewhat arbitrary, it should be easy for all reasonable people to agree that Socrates provides a better baseline than King Arthur. By establishing this as a reference class for our probability estimates, we significantly reduce the guesswork.
If this all sounds too complicated, Carrier will eagerly remind you that finding truth is never easy. Since the Bayesian method is merely a more precise and reliable reduction of all other methods, including your own feelings and intuition, any attempt to reject the method on the grounds that it’s too hard is a tacit admission that you’re not interested in truth.
This book is an invaluable tool for anyone with a serious interest in history. You will simply not find a more cogent, articulate, or exhaustive defense of scientific reasoning in a historical context. Dr. Richard Carrier has built a lighthouse of logic to guide all weary travelers on our endless quest to escape from ignorance.