"Proving History" by Richard C. Carrier

[BaalChatzaf]

"Proving History" by Richard C. Carrier 2012  Prometheus Books. This is a book show how Bayes Theorem is applied to historical queries.  Since the book is for historians it does deploy heavy duty mathematics.  However good mathematical references are given in the Appendix  and in the notes to Chapter 3.

Carrier  proves the following:  Bayes Theorem is the proper logical form for inductive arguments  In fact Bayes Theorem encapsulates the logic  of inferences made in physical empirical science.  

The particulars are an energetic  Bayesian deconstruction of arguments for the historicity of Jesus.  When Carrier gets done all there is is a pile of rubble.

It is good to see a  non-mathematical book do a thorough job of explaining what Bayes Theorem is and how it is used  in making inductive inferences. 

For those of you not familiar with Bayes Theorem please see https://en.wikipedia.org/wiki/Bayes'_theorem

The theorem itself:  P(A|B) = (P(B|A)*P(A))/P(B)  

P(A|B) is the probability of A  given that B is true.   It is required that P(B) > 0  where P(B) is the probability of B. 

 It is remarkable that the proper way of making  inferences when the premises are  uncertain is encapsulated in this rather simple formula.

 

[9thdoctor]

18 hours ago, BaalChatzaf said:

Since the book is for historians it does deploy heavy duty mathematics. 

You meant "doesn't".  And this was just the first book, the second is called On the Historicity of Jesus, and goes into that topic in much greater depth.  I found them both very worthwhile.  I actually started a thread about Carrier years ago that didn't attract much interest.

Here he is presenting Bayes Theorem for historians, an overview for the uninitiated:

BTW his analyses need to go through a lot more public critiquing.  Myself, I don't read Greek, so I'm not in a position to judge the way he handles the evidence.  He could be a total crank and I wouldn't know until another scholar takes him on and actually answers his arguments.  So I remain agnostic on the question.  Though I've always regarded it as a matter of probabilities, a context where capital C certainty can never be achieved. It's kind of like Global Warming "denial", the deniers (critics) are rejected out of hand by the (supposed) experts, and their arguments go unrefuted. 

[BaalChatzaf]

1 hour ago, 9thdoctor said:

You meant "doesn't".  And this was just the first book, the second is called On the Historicity of Jesus, and goes into that topic in much greater depth.  I found them both very worthwhile.  I actually started a thread about Carrier years ago that didn't attract much interest.

Here he is presenting Bayes Theorem for historians, an overview for the uninitiated:

BTW his analyses need to go through a lot more public critiquing.  Myself, I don't read Greek, so I'm not in a position to judge the way he handles the evidence.  He could be a total crank and I wouldn't know until another scholar takes him on and actually answers his arguments.  So I remain agnostic on the question.  Though I've always regarded it as a matter of probabilities, a context where capital C certainty can never be achieved. It's kind of like Global Warming "denial", the deniers (critics) are rejected out of hand by the (supposed) experts, and their arguments go unrefuted. 

Doesn't it is.  I did a typo.  Carrier is a first rate historian.  He has proved that Baye's theorem properly encapsulates not only scientific arguments  but historical arguments.  With history one is dealing with uncertain and incomplete data so probability must be used to come to conclusions.

Bayes Theorem is a proven mathematical result and it is rather easy to show.  It is using Bayes theorem that leads to complications.  

P(A|B) = P(A & B)/P(B)  by definition

P(B|A) P(A)  = P(A&B)/P(A) * P(A) = P(A & B) = P(A & B)/PB) * P(B) = P(A|B) * P(B)

so 

P(B|A) P(A)  = P(A|B) P(B)   hence

P(B|A) = P(A|B)*P(B)/P(A)    That is Bayes theorem.  Any  type of inductive argument that contradicts it must be invalid.

You really should read the book (which was published last year).  In it he corrects some errors he made in earlier work and shows the proof of the adequacy and necessity of Bayes Theorem in historical arguments.  This requires a lot of detailed analysis of historical material and historical hypothesis.  In a chapter about 80 pages long he pretty well demolishes all proofs from the believing historians who claim to have shown that it is probable that Jesus was an historical figure.  He quantifies the arguements and applies Bayes Theorem.  The details are in the book and I will not attempt to outline them here.

There is another book of interest: "Scientific Reasoning: The Bayesian Approach"  by Howson and Urbach.  The experimental tests of various physical hypotheses are analyzed using Bayes Theorem.   

The recent discovery of the Higgs Boson at CERN made heavy use of Bayes Theorem.  This may surprise you, but even accurate experiments in high energy physics  have error intervals that have to be integrated into the proofs or disproofs of hypotheses.   A lot of heavy duty statistics went into showing that the particles spotted in only a small subset of the experiments (particles with about 125 GeV  mass were with  high probability the Higgs Boson.  Please see 

http://hea-www.cfa.harvard.edu/astrostat/Stat310_1314/dvd_20140121.pdf

and

http://tonyohagan.co.uk/academic/pdf/HiggsBoson.pdf