Here's the relevant exchange re the hypothesis issue, where even taking Bruce's new and far more limited hypothesis into account, Billy points out the continuing invalidity of the approach:
Bruce E. Dale
on May 29, 2019 at 1:26 pm said:
Billy,
More on the calendar issue later, but the null hypothesis you propose is not the null hypothesis we actually deal with in the paper.
I wish you do us the courtesy of focusing on what we actually did in the paper, not what you think we should have done.
Our “null hypothesis” or the Bayesian prior we assumed, was that the Book of Mormon has nothing to do with ancient Indian cultures as Dr. Coe describes those cultures in The Maya.
Just set the Bayesian analysis aside for a while and focus on the correspondences without weighting them. Even in that limited case, it is obvious that the Book of Mormon has a great deal to do with ancient Indian cultures as described in The Maya.
We are going to go through those correspondences as long as the discussion continues or Interpreter is willing to host the discussion.
Once again, I invite you to read The Maya and the Book of Mormon…then make your own comparisons. And once again, I will happily buy a copy of The Maya for you.
Bruce
Billy Shears
on May 29, 2019 at 3:49 pm said:
Hi Bruce,
I’m happy to look at the specific points you raise, read the references in The Maya in detail, read the references in the Book of Mormon in context, consider your analysis, and then offer my own.
However, if you set up your analysis in a way that is invalid and insist that I approach the question in the way that you set it up, I’m not going to play ball.
Here is why.
In your paper, you said, “For a good introductory article to Bayesian statistics, see Wikipedia, s.v. ‘Bayes Theorem.'” According to that article, “If the events A1, A2,..., are mutually exclusive and exhaustive, i.e., one of them is certain to occur but no two can occur together, and we know their probabilities up to proportionality, then we can determine the proportionality constant by using the fact that their probabilities must add up to one.”
This point is crucial, because the formulas you are using implicitly assume that the two hypotheses are in fact mutually exclusive and exhaustive.
The way I am setting up the problem (A1 = “19th Century fictional origin” and A2 = “ancient Mesoamerican origin” at least approaches two mutually exclusive and exhaustive theories. In contrast, the way you are insisting this be set up (A1 = “Book of Mormon has nothing to do with ancient Indian cultures” and A2 = “the Book of Mormon is an authentic, factual record set in ancient Mesoamerica”) is neither mutually exclusive nor exhaustive: the book could (and does) have many superficial similarities with ancient Indian cultures, yet the totality of the evidence points strongly to 19th Century American origins.
We could find an arbitrarily high number of superficial similarities between the Book of Mormon and the Maya, but such comparisons have no bearing on whether the Book of Mormon is “an authentic, factual record set in ancient Mesoamerica.” If we want to use a Bayesian analysis to answer that question, it needs to be set up properly with two mutually exclusive and exhaustive hypotheses. The mere question “What is the probability Joseph Smith guessed X right?” is not a valid Bayesian likelihood ratio.
If you think I’m being unreasonable or discourteous by using valid Bayesian reasoning to address the question of whether “the Book of Mormon is an authentic, factual record set in ancient Mesoamerica” rather than using invalid reasoning to address the question of whether “the Book of Mormon has nothing to do with ancient Indian cultures,” then say the word and I’ll respectfully refrain from making any further comments.
Best,
Billy