The Effect of the Joseph Smith Papyri Being Found 1967-1968

[BeNotDeceived]

OK, let's see:

,,.
2. Smith created the Book of Mormon with help from Sidney Rigdon by plagiarizing from 5 books written before 1820 (View of the Hebrews, The Late War, The First Book of Napoleon, Manuscript Found and Captain Kidd).
............

Why is there any argument of the truthfulness of the church?

Believers can seize on specific instances of critics overstating their case and use that as a reason to disregard the rest which may have serious substantial issues.

You nailed that one!

[Philo Sofee]

Mr. Stak
Bayes theorem is a sausage. In some recipes it is the perfect sausage to use and in other recipes it is a poor choice. Carrier sells this sausage, but he doesn’t know how the sausage is made, and so his recommendations for recipes are going to eventually lead to a culinary disaster.

Philo
E.T. Jaynes recipes, for those who love a Bayesian feast, will lead to a culinary smorgasbord, so it’s all good.

Mr. Stak
Trying to use Bayesian reasoning to encapsulate the tools of reasoning in the historian’s craft is a project doomed to failure. This is necessarily so. I’ll endeavor to explain what I mean.

Philo
Simply incoherent. It’s already being used in history, as well as former history of problems frequentists using their imaginary ad hockeries could not solve, which Bayes does.

Mr. Stak
To do the things that Carrier wants to do requires him to use a non-trivial calculi to power the engines of bayesian computation. Each of the variables you have used in each of your formulas have something called “inductive content” that is expressed by the value you assign it. If you are doing what Carrier is advising, then each of those variables is in actuality a set of propositions about hypothesis, evidence, etc.

Philo
Bayes is used based on our current state of knowledge. Every time we accept something as true, probably true, maybe/maybe not, or even not likely, we are assigning values to what we already know about the world, and in relation to anything new we learn which may affect what we believe and doubt. As new knowledge shows up, new evidences amass, we update our knowledge. Otherwise we remain infants in knowing. This ability within humans is one of our most precious gifts we have. We are comparing ratios of probabilities of our information we have. Since we do not have all the information in the world, we upgrade wisely as we continue to learn instead of stopping at any way point which may be comfortable.

Mr. Stak
When you do the computations, what is actually happening is that each of those sets are being manipulated through Boolean operations.

Philo
Yes, thank goodness, as Cox’s theorem has demonstrated fundamentally proving the Boolean operations are the very best way to think in a rational manner and consistent manner based on the logical incorporation (using Bayes) of ever in-flowing new information.

Mr. Stak
Bayes really isn’t anything more than using deductive relationships to represent inductive strengths. For example just look at the Kolmogorov axioms, it is adapting probability to the structure of deductive reasoning.

Philo
Bayes is not, as E.T. Jaynes unerringly noted “the absolute status of an hypothesis embedded in the universe of all conceivable theories, but the plausibility of an hypothesis relative to a definite set of specified alternatives, that Bayesian inference determines.” (Probability Theory,” p. 310.) He further notes - “The functional use of induction in science is not to tell us what predictions must be true, but rather, what predictions are most strongly indicated by our present hypotheses and our present information.” (p. 310)

The weakness of the Kolmogorov axioms is that his system makes no reference to the notion of conditional probabilities. Yet obviously, as Jaynes notes in his discussion of the Kolmogorov view, all probabilities to the real world are necessarily conditional on the information at hand. (p. 654).

From the stand point of logic the product rule (and therefore Bayes’ theorem) expresses both the associative and commutative properties of Boolean algebra. “That is what gives us that greater freedom of action in calculations… the complete freedom to move propositions back and forth between the left and right hand sides of our probability symbols in any way permitted by the product and sum rules. This is a superb computational device - and by far the most powerful tool of scientific inference…” yet completely missed by the Kolmogorov system of probability. (p. 654)

Mr. Stak
Now the sets that make up that inductive content are not exactly uniform in their complexity. Imagine the complexity of one hypothesis being that of a straight line: y = ax + b. It has just two adjustable parameters and all the possible computations are contained in a set. Now imagine the complexity of another hypothesis being that of a parabola: y =ax^2 + bx + c. Now you have to deal with three adjustable parameters and thus the content of the set just expanded exponentially. This will create problems down the line.

Philo
Which Bayes theorem is designed to take care of through multiple hypothesis testing, taking things one step at a time. It does not deal with all infinite possibilities, but limits its use to finite sets and parameters. Problems being created down the line is no reason to ignore Bayes, but to utilize it as well as it is possible within the information content of our knowledge.

Mr. Stak
Ray Solomonoff knew this all too well and that's why he always insisted on the use of the most algorithmically simple hypothesis you can skate by with. The problem with trying to model historical analysis is that you can’t use a simple hypothesis. The sets you are dealing with are impossibly large.

Philo
Not when it comes to whether I am to believe something or not based on what I know. I can always enlarge my knowledge and include ever larger circles of sets of information. Nothing in Bayes requires I take all information sets in its most gigantic arena all at once. You can eat an entire elephant by yourself, one bite at a time. Finite sets seriously gigantically large as shown by Jaynes, are how Bayes can handle them. I don't have an impossibly large set to deal with in testing the historical claim that Joseph Smith translated Egyptian papyri correctly at all. I have all the evidence and all the background knowledge from 1835-2021 - and its manageable in order for me to objectively see if it is probable that he translated it correctly. He didn't, based on the information and evidence we possess right now.

Mr. Stak
You know when you “update” in Bayes? Well what you don’t see in the algebra but is taking place is called disjunctive refining. In mathematical logic it is like adding the word “or” to a well formed formulae (or in Carrier’s case a proposition) and then introducing a new well formed formulae (or proposition). This is supposed to capture the new evidence coming in and signify a rival hypothesis.

Philo
So new information can be used to update what we know. It’s not on a rival hypothesis on a belief I may have, it is further information which I can use to update whether I should continue in the trajectory of what I think. If we are testing two hypotheses, then the Bayes ratio sets it up so we can legitimately compare them with the information we have and evidence for either one, and update accordingly. That is entirely the rational thing to do. It's conditional based on the information we have and understand right now. It's not a final probability to absolute truth, and never has been, nor has it ever, ever, ever, ever been defined in such a silly way.

Mr. Stak
When you start disjunctively refining these massively complex sets you’ve built to house the entirety of historical knowledge, you quickly reach a point where disjunctive refining no longer captures new inductive content. Your new evidence no longer makes any difference whatsoever. Why does this happen? It is a consequence of Boolean algebra, it has to maintain something called “symmetry” or to use a more precise term “labeling invariance”. Basically this means all those propositions in your set have to be arranged in a homogenous fashion and disjunctive refinement is going to throw it all out of whack.

Philo
And if Jaynes knows what he’s talking about there are cures and correctives for this. It is precisely the Boolean algebra which gives the proper ways of utilizing the information within a Bayesian juncture of ratios which, in point of very fact, keep one consistent and rational.

Mr. Stak
In other words, Bayesian probability eventually has to deal with the fact that it lacks logical completeness. All the noise about how to handle “the priors” doesn’t really come from a concern about external sources masquerading as inductive strength in the guise “bias”, they are created because the deductive symmetry that helps define the relationships between all those variables is going to break down and any new evidence will simply be irrelevant in terms of mechanics, so you can use priors to help stave that off. For a little bit.

Philo
Bayes has never, ever, ever been about logical completeness. Holy cow man… bias is already in every single decision you make in life, Bayes does not hide it, it brings it out in order for the ratios, evidences, and background knowledge to limit it’s running away amok. Priors don’t stave off anything. Evidence and background knowledge and new information does that. Holy cow man… Priors are simply what you know at a given moment, which are not the entire enterprise of updating our information. Priors are the beginning, not the end of a limited conditional process for probabilistic rational thinking. No one in all of the literature has ever said Bayes is all about getting to logical completeness. This is entirely a red herring about what Bayes is all about. One never arrives at certainty with Bayes, but at a good probable estimate of the strength of one’s belief, conditional on what is known. Conditional. Not certain. Conditional, not complete. Conditional.

Mr. Stak
What I’m telling you isn’t new.

Philo
It certainly isn’t accurate either.

Mr. Stak
It has been acknowledged and accepted since the first half of the 20th century.

Philo
And has been refuted in the 21st century. Why stay in old grooves when Bayes is especially designed to upgrade our knowledge? It is fundamentally not accepted by all Bayesians. We can’t even recognize your definition of what Bayes is supposed to be doing, and that is simple fundamental first grade stuff to know.

Mr. Stak
You will never be able to use probabilistic reasoning to meaningfully capture historical reasoning because the complexity of sets will fail you long before you can get there. It isn’t even up for debate. The matter is settled in the strongest possible way.

Philo
Pure Posh. Nate Silver absolutely thoroughly refutes this assertion (with no evidence) in his stellar book “The Signal and the Noise.” E. T. Jaynes makes hash of all your assertion here entirely throughout his 700 page magnum opus “Probability Theory.” John Earman “Bayes or Bust?” certainly doesn’t agree with how you define and describe Bayes either. Now had you said by frequentists, then you might be on stronger grounds, but there are thousands of Bayesians who certainly would disagree.
Luc Bovens and Stephan Hartmann, “Bayesian Epistemology,” note how skewed your entire background knowledge of what Bayes is supposed to be doing, as opposed to what it actually is all about - “The probability of the consequence given that the hypothesis is true is greater than the probability of the consequence given that the hypothesis is false.” (p. 90)
Notice, there is nothing concerning logical certainty or completeness. How you think this is what the goal of Bayes is is truly beyond me. Obviously you aren’t looking into the Bayesian literature much, because there is not one book on it which says what you are here. No book says we have to go back infinitely testing everything single detail of history in order to come to a justified objective probability about whether we have the right belief in something or not. To test a historical probability of whether Joseph Smith correctly translated the papyri is not tantamount to me having to go back into the pre-dawn of Egypt, learn everything about every item of clothing ever donned on by all people, test their eye color and find out what they have eaten through the millenia. That is palpably ridiculous to imagine. I stand exactly by my probability on why I disbelieve based on what I know, and the evidence we have. Show me specifically how my example is wrong then. Show me other evidence which refutes my own probability, and then show me what the actual probability ought to be based on everything I have presented then. Let's talk this through with Bayesian thinking. I am totally game with that.

Mr. Stak
This is why Carrier is hocking his books in the world of counter-apologetics and Biblical Studies.

Philo
Ad homimem, hence irrelevant, to his actual use which I see no refutation directly from you. Mind you, I am not saying Bayes is infallible, not at all. No one ever has or does. But there are many who say it is a sensational tool in ever widening fields of use and practicality. It is rather sensation to use in historical issues concerning Mormon claims. If that isn’t obvious, then I sincerely have to ask what you actually know about how Bayes works with hypotheses, because I am finding it works very well in historical matters of claims in Mormonism in order to see if I ought to be believing them or not, again, using it with evidence and background knowledge of how the world works and what information I have at hand.

[Jersey Girl]

This is why Carrier is hocking his books in the world of counter-apologetics and Biblical Studies.

hawking.

Sorry.

[Shulem]

5. Smith married 33 women, several being teenage girls.

Non sequitur!

That's what prophets do, they F U C K women, and lots of 'em. Horny Jacob f u c k e d four women to come up with the House of Israel which is built upon the foundation of holy-polygamy. Jacob had a ten inch cock! Don't you know that? Holy women took turns to enter the holy tent in order to conceive and get their next baby. Just read the Bible, it's all in there! Mixed marriages too! Abraham f u c k e d Egyptian women. And Hagar was a whore.

6. Smith sent men on missions and then propositioned their wives while they were gone.

And prophets do that too! They have needs in order to be fulfilled.

[Shulem]

"Can you not see that I'm a woman?"

We've come a long way, baby.

Facsimile 3: Return of the False Prophet or Mormonism Disproved Again!

[DrStakhanovite]

Hi Philo,

So you and I don’t have to write increasingly long posts back and forth, I thought it would better facilitate our chats if we spent some time hashing out just one issue at a time.

When you do the computations, what is actually happening is that each of those sets are being manipulated through Boolean operations. Bayes really isn’t anything more than using deductive relationships to represent inductive strengths. For example just look at the Kolmogorov axioms, it is adapting probability to the structure of deductive reasoning.

Bayes is not, as E.T. Jaynes unerringly noted “the absolute status of an hypothesis embedded in the universe of all conceivable theories, but the plausibility of an hypothesis relative to a definite set of specified alternatives, that Bayesian inference determines.” (Probability Theory,” p. 310.) He further notes - “The functional use of induction in science is not to tell us what predictions must be true, but rather, what predictions are most strongly indicated by our present hypotheses and our present information.” (p. 310)

The weakness of the Kolmogorov axioms is that his system makes no reference to the notion of conditional probabilities. Yet obviously, as Jaynes notes in his discussion of the Kolmogorov view, all probabilities to the real world are necessarily conditional on the information at hand. (p. 654).

From the stand point of logic the product rule (and therefore Bayes’ theorem) expresses both the associative and commutative properties of Boolean algebra. “That is what gives us that greater freedom of action in calculations… the complete freedom to move propositions back and forth between the left and right hand sides of our probability symbols in any way permitted by the product and sum rules. This is a superb computational device - and by far the most powerful tool of scientific inference…” yet completely missed by the Kolmogorov system of probability. (p. 654)

The underlined section of what I said was quoted by you and the above three paragraphs written by you is a response to that. At least, that is how I took it. I’m having difficulty connecting what I said with your response.

For example, I mentioned the Kolmogorov axioms being examples of probability conforming to the structure of deductive reasoning, but your comments about the Kolmogorov axioms seem unrelated to that. Could you say more about how what you wrote interacts with what I’m saying? For example, how does “The weakness of the Kolmogorov axioms is that his system makes no reference to the notion of conditional probabilities.” relate to my observation about Kolmogorov axioms?

Also I’m having an issue with Jaynes. I have his ‘Probability Theory: The Logic of Science’ open to page 310, but it is in the middle of a section titled ‘Logic vs Propensity’ and I can’t seem to find what you are quoting from. I’m confident it is just an issue of pagination, are you using a digital copy perhaps?

[DrStakhanovite]

hawking.

Sorry.

I was going for the spitting imagery

[Jersey Girl]

hawking.

Sorry.

I was going for the spitting imagery

Oh? Now you're just getting disgusting.

[Philo Sofee]

Hi Philo,

So you and I don’t have to write increasingly long posts back and forth, I thought it would better facilitate our chats if we spent some time hashing out just one issue at a time.

When you do the computations, what is actually happening is that each of those sets are being manipulated through Boolean operations. Bayes really isn’t anything more than using deductive relationships to represent inductive strengths. For example just look at the Kolmogorov axioms, it is adapting probability to the structure of deductive reasoning.

Bayes is not, as E.T. Jaynes unerringly noted “the absolute status of an hypothesis embedded in the universe of all conceivable theories, but the plausibility of an hypothesis relative to a definite set of specified alternatives, that Bayesian inference determines.” (Probability Theory,” p. 310.) He further notes - “The functional use of induction in science is not to tell us what predictions must be true, but rather, what predictions are most strongly indicated by our present hypotheses and our present information.” (p. 310)

The weakness of the Kolmogorov axioms is that his system makes no reference to the notion of conditional probabilities. Yet obviously, as Jaynes notes in his discussion of the Kolmogorov view, all probabilities to the real world are necessarily conditional on the information at hand. (p. 654).

From the stand point of logic the product rule (and therefore Bayes’ theorem) expresses both the associative and commutative properties of Boolean algebra. “That is what gives us that greater freedom of action in calculations… the complete freedom to move propositions back and forth between the left and right hand sides of our probability symbols in any way permitted by the product and sum rules. This is a superb computational device - and by far the most powerful tool of scientific inference…” yet completely missed by the Kolmogorov system of probability. (p. 654)

The underlined section of what I said was quoted by you and the above three paragraphs written by you is a response to that. At least, that is how I took it. I’m having difficulty connecting what I said with your response.

For example, I mentioned the Kolmogorov axioms being examples of probability conforming to the structure of deductive reasoning, but your comments about the Kolmogorov axioms seem unrelated to that. Could you say more about how what you wrote interacts with what I’m saying? For example, how does “The weakness of the Kolmogorov axioms is that his system makes no reference to the notion of conditional probabilities.” relate to my observation about Kolmogorov axioms?

Also I’m having an issue with Jaynes. I have his ‘Probability Theory: The Logic of Science’ open to page 310, but it is in the middle of a section titled ‘Logic vs Propensity’ and I can’t seem to find what you are quoting from. I’m confident it is just an issue of pagination, are you using a digital copy perhaps?

Yeah I probably was too hasty, it's in there, I shall look when I get back to work. I am much more interested in seeing why you imagine one cannot upgrade epistemic beliefs on historical issues using Bayes. That's precisely where I find Bayesian thinking at its truly best and strongest. As Jaynes has described so very well how thanks to R. T. Cox and George Polya and others we can help keep our thinking consistent and rational, and shows with so many varied subjects how that works. In my personal applying it to my beliefs on claims in Mormonism and how it has changed the way I ask myself questions I find it truly eye opening! I have new angles to approach apologetic claims now also, but most especially, in asking myself why did I believe this or that then, but not now? What exactly has changed for me.
Kolmogorov axioms and all that is fine to discuss, but irrelevant to how Bayes is helping me thinking through my thinking and beliefs and now my doubts. That is what is most intriguing to me.

[Physics Guy]

Really just a side remark, not particularly in response to other posts here: about my own reading of E.T. Jaynes.

Most of my research is in the large and general sub-field of physics called "statistical mechanics", or at least related to it. I think that's largely because I never took a proper course in the subject during my studies. So I never got brainwashed, as most physics students do, into accepting that everything is all clearly understood in this field, nothing to see here move along. In fact it's really the soft underbelly of physics where a lot of stuff is not well understood at all—so the potential opportunity to discover something new is still high.

The brainwashing really seems to be a thing with this particular subject. Because its foundations are actually weak, lots of people seem to want to insist that they're not weak, and they've developed apologetic arguments to try to show that the stuff that seems to work, empirically, is indeed fully understood as logically necessary truth.

E.T. Jaynes was one of these brainwashers, I eventually concluded. He was a physics professor and the use of Bayesian inference that he championed most was in statistical mechanics. An introductory paper by him on the subject, in a "special topics" course in honours year, was the closest I came to formally studying "stat mech".

It was appealing. He made everything look clear and simple. But I later came to believe that it was a horrible approach to the subject which merely begged the real questions—or swept them under the rug.