Analytics

What is your theory about how the 1837 poem and 1850 novel had access to the exceedingly unlikely phrase "save it were"? Is this proof that the respective authors had read the Book of Mormon and picked it up from there? Or is it evidence that these literary works are actually ancient documents translated with the help of English ghosts? Or is it evidence that 19th century New Englanders had access to this phrasing through spoken language, language that wasn't captured in your database, reading older books, or otherwise recreating features of the language that make sense and sound good even though they weren’t frequently used? I would suggest there is a naturalistic explanation for why the 1837 poem and Nathaniel Hawthorne had access to this phrase, despite the gaps of its appearances in your database. Is it really that crazy to think the author of the Book of Mormon could come up with it in the same way? This is your area of expertise and not mine, but I would expect the early 18th-century English of New England (expanded to include upstate New York) would be closer to English English than to the English spoken in, say, Georgia. And I'll go out on a limb here. My hypothesis is that one of the reasons "save it were" is *rare* in your database is because this usage was generally limited to New England. If I'm right, then I'd expect the 1837 poem to be from there and not, say, from South Carolina. Am I right? Where was the 1837 poet from? Could you explain why using "save it were" over 70 times is evidence of ancient American authenticity? If this were in the lexicon of the 19th-century author and it sounded old to his ears, why wouldn't he use it over and over for the same reason he used "and it came to pass" over and over?

Sigh. You fundamentally misunderstand the nuances of Bayesian reasoning and why I keep lecturing about it. Bayesian reasoning actually has two components. First and most importantly, it informs us as to what constitutes evidence and why. As Steven Pinker explains it: "Bayes' rule or Bayes' theorem is the law of probability governing the strength of evidence--the rule saying how much to revise our probabilities (change our minds) when we learn a new fact or observe new evidence." Pinker, Steven. Rationality: What It Is, Why It Seems Scarce, Why It Matters (p. 149). Kindle Edition. This is why I've been repeating the steps of Bayesian analysis over and over. You can't correctly interpret "the strength of evidence" without at least implicitly formulating the questions in the correct Bayesian structure. The correct way to understand the strength of evidence--and really what evidence even is -- comes down to two questions, asked either explicitly or implicitly: What is the probability of seeing the basket of evidence if the theory is true? What is the probability of seeing the basket of evidence if the theory is false? To correctly interpret the evidence, you need to consider both in relation to each other. If you don't, you can dramatically misrepresent what the evidence implies. Note that so far I have not invoked my prior at all--I've only talked about how to weigh the strength of the evidence. A priori beliefs are a different discussion, and surprisingly it's logically required: you cannot make a decision about the likelihood of something without having an a priori belief about it. I start with a low prior for the BoM being true based on general scientific literacy, and then lower it significantly based on the evidence of what the text actually contains. @Ryan Dahle's assertions aside, I believe an unbiased evaluation of the evidence clearly indicates it is 19th century biblical fan fiction--not an accurate translation of an authentic ancient manuscript. Newly considered evidence should generally be expected to further support already well-supported a priori beliefs, and in this case it does. As Nate Silver said, "One property of Bayes's theorem, in fact, is that our beliefs should converge toward one another--and toward the truth--as we are presented with more evidence over time." Silver, Nate. The Signal and the Noise: Why So Many Predictions Fail--but Some Don't (p. 259). Kindle Edition.

Taking your hypothetical seriously, if you pulled out a shrink-ray gun and demonstrated that it worked on your mailbox, my first thought would be that there was a hidden camera and I was secretly being recorded on The Carbonaro Effect. If that possibility were ruled out, I wouldn’t start by asking why you were abducted, why you were given the gun, or how the aliens built it. Instead, I would want to examine the evidence. Specifically, I’d want a qualified expert to evaluate the device under controlled conditions. I’d ask a physicist to determine what the “shrink ray gun” actually does in the real world. Does it compress matter? Alter particle structure? Cause matter to disappear? If so, where does it go? Once we understood the mechanism, I would ask whether the device appears consistent with known materials and technology available on Earth. If experts concluded that the device genuinely performs phenomena that cannot plausibly be produced by known terrestrial technology, that would be extremely strong evidence, and I would update my credences accordingly. But to evaluate the alien-abduction claim coherently, I’d need to consider: • the prior probability that you were abducted by aliens • the probability of observing such a device if aliens really had abducted you • the probability of observing such a device if aliens had not abducted you If the device strongly disfavors terrestrial explanations, the likelihood ratio would heavily favor the alien-abduction hypothesis. But if you said something like, "The alien really wanted me to prove to the world that he was real, so he commanded me to show the shrink-ray gun to eleven close friends who then jointly signed a statement saying it was real—and then the alien who wanted to prove his existence took the gun away before any qualified physicist could examine it," I would be far less impressed.

This concept that there are multiple “types” of English translations that would be made by the power of God and that one of them involves a heterogeneous array of speakers with English dialects that span from the 1500s through the 1800s is an example of Texas sharpshooting because this idea didn’t come about until after the evidence came out. Originally, there were only two types of translations: correct translations and incorrect translations. Implausible compared to what?

Let’s back up and put this into perspective. My Bayesian reasoning has three stages, and the order matters. Stage 1 is the prior. The base rate of genuine miracles in all of recorded human history is extraordinarily low. This isn't dogma. It's an empirical observation. As I've repeatedly hammered, multiple lines of science converge on the point that not only is the supernatural not observed, it isn't even possible (and in this context "supernatural" entails "spirit matter" and the alleged higher laws of nature that Mormonism posits). My prior probability for the BoM being authentically what it claims is therefore very low before I examine a single piece of BoM-specific evidence. Stage 2 is what the Book of Mormon actually is. Before asking who wrote it or how, I look at the text itself. This is the most important question and the one you spend the least time on. The text addresses 19th century theological debates. It contains material from Deutero-Isaiah--chapters written centuries after Lehi would have left Jerusalem. It has no credible archaeological, genetic, or linguistic corroboration as an authentic ancient record. Rather, it is full of anachronisms. The text itself is primary evidence, and it looks like 19th century composition. Stage 3 is mechanism. Only after establishing what the BoM is do I ask who wrote it and how. On this I am and always have been agnostic, and appropriately so. We don't have enough evidence to reconstruct the precise mechanism. You asked me to speculate and I did. But agnosticism about mechanism is not weakness. It's epistemic honesty. You are conflating two separate questions: whether the BoM is authentic versus how it was produced. I am highly confident on the first and appropriately uncertain on the second. Those are not contradictory positions. Now on your Bayesian critique: you say I'm not being rigorously Bayesian because I don't show explicit priors and likelihood ratios. But it is okay to be informal and not quantify terms like "unlikely" and "very unlikely." The key Bayesian insight in this context isn't that we need to make up probabilities and do math. The key insight is that we need to ask the right questions: What is the a priori probability? What is the probability of seeing the evidence under one theory? What is the probability of seeing the evidence under the competing theory? It's fine to discuss these probabilities qualitatively, but it is a logical fallacy to evaluate the probability of the BoM being "true" without addressing all three components. On the witnesses: they were chosen by Joseph Smith and their testimonies were first drafted by hin and then signed. That is not independent testimony. It's managed testimony. A judge probably wouldn't admit evidence managed this way, and if she did, she wouldn't give it much weight. I don't either, for the same reasons. It isn’t about arbitrary rules of evidence. It’s about the evidence being suspicious because it is managed. On the plates: performative evidence that was deliberately kept from independent examination and then conveniently removed is the structure of a magic show, not an authentic historical artifact. This pattern continues for most of the evidence we have about how it was produced. The "evidence" was tightly managed with the objective of proving it is miraculous while going to extraordinary lengths to prevent genuine authentication. When you don't have enough evidence to know exactly how something was done, the honest thing to do is say "I don't know." That’s why I hold the position that I do.

We need to remember that likelihood is a relative term. Say I flipped a coin 100 times and got the following string of values: T,T,H,T,H,T,T,T,H,T,H,T,H,T,T,H,H,H,H,H,T,T,T,T,H,T,T,H,T,T,T,T,H,T,H,H,H,H,T,H,H,H,H,T,H,H,T,H,T,H,T,T,H,H,H,H,T,H,H,T,T,T,H,T,T,H,T,T,T,T,T,H,T,H,T,T,H,H,T,T,H,T,H,T,H,T,H,T,H,T,H,T,T,H,T,T,T,H,H,T Under the hypothesis that the probability of getting a heads is exactly 50%, the likelihood of getting that specific set of results with 100 flips is 7.89x10^(-31). In absolute terms, that is extraordinarily unlikely. But it is also extraordinarily unlikely that any other probability of getting a heads would result in that data. If we wanted to evaluate the likelihood that the coin is fair based on the data, we can’t look at the likelihood in absolute terms--we need to evaluate it compared to the likelihood of other hypotheses. So yea. The probability of “save it were” showing up in a 19th century text that is trying to use the verbal sacred language of the time and place is low in absolute terms. But virtually everything involving stochastic processes is low in absolute terms.To determine what bearing this has on the probability that the Book of Mormon is true, you need to show how and why “save it were” is more likely under the competing hypothesis. You seem to be claiming that using “save it were” is more likely under the “it’s true” hypothesis, but you didn’t determine this until after you saw it there. That is classic Texas sharpshooting. I tip my hat to Skousen and Carmack for acknowledging this, and I have no problem whatsoever with them describing the language of the Book of Mormon. If they want to assume the BoM has a divine origin, then explaining this data seems to require the “tightly controlled transcription of a loosely translated document created by a committee of angels who had heterogeneous English dialects that spanned over 300 years,” or whatever their specific hypothesis is. But it is a different question to ask what this implies about the likelihood of the Book of Mormon being authentic. If you define the authentic hypothesis as "whatever divine process produced exactly the text we observe," then you are Texas sharpshooting, not seriously evaluating the likelihood of authenticity. A mixed linguistic profile spanning three centuries of English is precisely what you'd expect from a 19th century author steeped in biblical register and trying to sound biblical. And that’s precisely what demands the most creative explaining under divine translation. I disagree, but we can save this for another day.

In context, I offered two items of support. As an example, "save it were” is a phrase that, if I understand the argument correctly, Joseph Smith in all likelihood wouldn’t have used because that specific phrase isn’t in the Bible and isn’t in the body of digitized 19th century literature that Carmack searched. However, there are two reasons to believe that the phrase was in his lexicon. First and most importantly, everybody understood that wording--it was their language. Second, he used it. We already know from other evidence that the BoM is of 19th century origin. The Book of Mormon is part of the enormous limited dataset we have of 19th century usage. And equally importantly, we have an eminently plausible theory of how “save it were” entered his lexicon. After all, his lexicon is a descendant of EModE. Spoken english is different from written English, and Joseph Smith’s particular dialect of oral sacred English could have retained elements of EModE that weren’t written in the limited datasets that Carmack looked at. And the fact that it is there is strong evidence that this is in fact what happened. I’m not making a far-fetched claim here. I’m following the data. To be clear, this isn’t evidence in favor of 19th century authorship. Rather, it’s showing how this is not a silver bullet against a naturalistic explanation. "Enormous" is relative. Written corpora don't capture spoken language, and the sacred oral vernacular of Bible-reading frontier communities isn't well-represented in the databases Carmack searched. Absence from the databases Carmack searched doesn’t prove absence from the living religious speech of Joseph Smith's milieu. We are evaluating the claim “what does the evidence say about the likelihood that the Book of Mormon is true?” The (only) logically coherent way to do this is with Bayesian reasoning. To do this, we need to do the following: Create at least two broad, non-overlapping, well-defined theories of Book of Mormon origins that we are evaluating. Select an a priori likelihood of each theory being true Collect the evidence Determine the likelihood of seeing the totality of the evidence under each scenario Do the math Your argument is basically, “we don’t know anything about the supernatural, so we have no basis for evaluating the likelihood of the evidence under the supernatural hypothesis.” If that is your position, then fine--we don’t have a well-defined theory to evaluate and no analysis should be done or can be done on the “is the Book true” question. Stated in a non-Bayesian framework, this is functionally the same thing as making an unfalsifiable claim. An unfalsifiable claim really isn’t claiming anything at all. But if you want to say the data is overwhelming against the theory that the Book of Mormon was written in the 19th century, you must propose an alternative theory that is more consistent with the data than the one you are arguing against. It depends on which supernatural theory we’re talking about. The supernatural theory you seem to favor has something to do with a ghost committee of individuals with different English dialects spanning from the 1500s to the 1800s, providing some sort of text that is primarily 19th-century Bible fan fiction, but also may or may not have a few elements loosely based on something that happened in Mesoamerica--it’s impossible to know the specifics because we know so little about God and angels. You can make that your theory if you want to, but when you carefully draw your theory around the bullet holes after you shoot the gun, an honest a priori probability gets commensurately smaller.

That is an assumption that in all likelihood isn’t true. To prove that this assumption is true, you’d need to comprehensively provide transcripts of everything that was ever said or read in his environment. We don’t have that. You are fundamentally misunderstanding the Bayesian approach. The correct approach is to address the totality of the evidence under two scenarios: What are the chances this basket of evidence would show up if the Book of Mormon were true? What are the chances this basket of evidence would show up if the Book of Mormon were false? EModE linguistic forms are highly unlikely under both scenarios. Because of that, it isn’t strong evidence of authenticity either way. I want to be careful here because while I know a bit about statistics (e.g. I know what the Yates Chi-squared test is), I haven’t read his research--I’ve only read a few things about it. On the one hand, “save it were” might not have been the most common way to say “unless” in the 1820’s, but everyone knew exactly what it meant, and it’s eminently plausible--even likely--that this phrasing was not completely extinct in Joseph Smith’s universe--that is how they knew what it meant. On the other hand, the BoM contains words that didn’t exist in the EModE period (e.g. “transpire”, “derangement”), and religious debates that hadn’t started yet (e.g. things like infant baptism, trinitarian v. non-trinitarian theology, predestination vs. free will, anti-masonry rhetoric) all fit in Joseph Smith’s time and place and not in the EModE period. What is more likely? Somebody in the 19th century writing a 19th century book in KJV English had mental access to a few remnants of EModE that haven’t shown up in other 19th century documents? Or whatever Carmack’s theory is about a ghost committee that spoke multiple dialects of English depending upon the verse? The way I account for it seems a lot more likely and clean than the way he does.

The “more recent findings” don’t alter the validity of Wright’s analysis. Does champatsch have a theory on why Deutero-Isaiah is in the BoM? As background: How do you account for Isaiah chapters in the Book of Mormon that weren’t written until long after Lehi left Jerusalem? It isn’t a matter of hypothetical precedent. It’s about evaluating the claims of the Book of Mormon. We don’t need to speculate about what God would do in this or that hypothetical situation. We can evaluate the Book of Mormon’s own claims. Have you read 1 Nephi 13? The Bible was corrupted by the great and abominable church. That is why we have the Book of Mormon, including the Isaiah chapters. It is also why Joseph Smith saw the need to create a new “translation” of the Bible. I’m not evaluating the hypothesis that God would do whatever random thing in whatever random way. I’m evaluating the Book of Mormon on its own terms.

Fair enough. I'm generally agnostic about the specific details of the mechanism of production, but since you've asked repeatedly, here are the scenarios I find most plausible, offered as my best current assessment. On the text: Joseph Smith is the most likely primary author, possibly with help from people in his immediate circle who were deeply steeped in biblical language. I can imagine his mother Lucy and brother Hyrum being involved. They had over five years between the time they said the book was forthcoming until the dictation to scribes began. The "tight translation" elements Carmack identifies are consistent with someone who had deeply internalized archaic KJV language and was reading from a pre-prepared manuscript, first behind the curtain during the 116-page period, then from the hat. The EModE patterns don't require a miracle. They require immersion in sacred language traditions, which Joseph's family had in abundance. On the plates: some physical object probably existed. It was never subjected to independent examination, was deliberately concealed, and then conveniently removed before anyone outside the inner circle could evaluate it. This would have been constructed during the 5-year period before they were shown to witnesses. On the witnesses: I think the most plausible explanation combines sincere belief in the project with solemn religious oaths to stand by what they signed. These were not dispassionate observers. They were family members, financial stakeholders, and true believers in a sacred undertaking. Several maintained their testimonies even after leaving the church, which I take as evidence of genuine belief in the original project, not necessarily evidence that the production was exactly what it was claimed to be. People can sincerely believe in something while using guile (e.g. missionaries who are told to find their testimony by telling people they know the Church is true). The witnesses may have believed they were serving God's purposes even while participating in a production that was partly staged. Moroni 7:16 explains their motivation--even if they weren’t being totally transparent and guileless, they were still persuading people to believe in Christ and do good; that is the yardstick for whether they were doing God’s work--not whether they were telling the truth, the whole truth, and nothing but the truth. On Space Aliens and demons: I never sincerely posited these as explanations. My point was logical: if the production was so extraordinary that naturalistic explanations are ruled out entirely, then all supernatural explanations need to be considered. And when choosing among supernatural explanations, it still matters that the text reads like 19th century biblical fan fiction rather than an authentic ancient record. But Space Aliens and demons are extraordinarily unlikely explanations and I never meant to suggest otherwise. On Penn & Teller: As a more on-point example, between 1964 and 2015, hundreds of people applied to win James Randi’s million-dollar prize for having purported supernatural abilities. All claimed real magic. Some were probably sincere. None could prove they had real powers under controlled conditions. Claiming supernatural powers is irrelevant to the point; I don’t need to know how the trick to know there is one.

My understanding is that your theory is that the translation process was very tight, to the extent where if the underlying reformed Egyptian had different frequencies of subordinating conjunctions after certain words, those patterns would be manifest in the translated text. Is that basically right? Could we use the Isaiah chapters to test this theory? If the Isaiah chapters were really a tight translation of a pristine Isaiah manuscript that was written on metal plates before 600 B.C., what would we expect to see? Do we see that? Here you don’t seem to be arguing that the Isaiah chapters are a better translation of Isaiah's actual words. Rather, you seem to be arguing that even though it is clearly derived from the KJV, it was done so in a miraculous way. This gets back to my point about the logical order of the analysis, and to the need to be clear about what the competing theories are. Is the theory that the Book of Mormon was produced in a miraculous way? Or is the theory that it is an accurate translation of an authentic ancient manuscript? Evidence of the first thing isn’t necessarily evidence of the second.

Hi @smac97, I appreciate your patience in my response. I’ve had a crazy couple of weeks IRL. Trying to respond to your main points as I understand them, you frame this as though I owe you a complete theory of the crime before I'm permitted to conclude that a crime occurred. That's not how epistemology works, and it's not how Bayesian reasoning works. The correct order of questions matters enormously, and you have it exactly backwards. First: What is the prior? The BoM being a genuine ancient record translated by divine power is an extraordinary claim--extraordinary precisely because it requires the causal mechanisms that Effective Quantum Field Theory rules out. Revelation. Spirits communicating information to a physical brain. Angelic delivery of gold plates. The prior probability of this hypothesis is therefore very low. Not zero, but very low. This is my point with the implications of physics that Carroll has articulated, and the exact same conclusion is robustly reached from multiple lines of science using completely independent methodologies. This is where any honest Bayesian analysis begins, before examining a single piece of evidence. Second: What IS the Book of Mormon? When you read it, this question answers itself. It is saturated with 19th century American theological concerns--evangelical Protestant disputes, anti-Masonry, the question of Native American origins. It reads like King James English filtered through frontier American religious sensibility, not like an authentic ancient document. It contains systematic anachronisms for which there is zero archaeological corroboration. David P. Wright’s analysis of the Isaiah chapters (Dr. Wright is a full professor at Brandeis--arguably America’s most prestigious Jewish university--where he specializes in teaching the Hebrew Old Testament at the graduate level. His exhaustive analysis of the Isaiah chapters has not been refuted--they were based on copying and reworking the KJV English translation--not a fresh translation of an authentic ancient manuscript). The DNA evidence provides completely independent confirmation that the population history the text describes didn't happen. The text accounts for itself. On the witnesses and plates: You want these "accounted for." The correct Bayesian question is: how likely is this evidence under each hypothesis? Under the authentic hypothesis, you'd expect witnesses. Under the "19th century American religious movement with a charismatic founder" hypothesis, you'd also expect witnesses--fervent, socially embedded, financially and spiritually invested witnesses whose accounts evolved over time and explicitly included "spiritual" rather than physical sight in several cases. Which is exactly what we have. Perhaps you’d expect witnesses more under the authentic hypothesis, but not necessarily, and certainly not enough to put a dent in the evidence against authenticity. The evidence of the plates is thin and entirely testimonial--and not innocently so. The alleged plates weren't lost in a fire or destroyed by accident. They were deliberately hidden from mainstream view and then vanished in a deliberate and miraculous way. Consider what that looks like from an evidentiary standpoint. Imagine you were involved in complex litigation with extraordinarily important physical evidence. You show that evidence exclusively to eleven of your client's closest friends and family members, obtain brief signed statements from them, and then deliberately destroy the evidence before the other side can examine it. No judge would treat that witness testimony as strong corroboration. Every judge would treat the deliberate destruction of the underlying evidence as deeply suspicious--suspicious enough to substantially discount the testimony built around it. You don't need a theory of why the witnesses signed the statements they did to draw that inference. The destruction itself does the work. One question for you: do you have a detailed, evidence-supported hypothesis for how God and/or Moroni made the plates disappear? If you don't need that hypothesis to believe the plates are a real ancient artifact, why do I need a detailed hypothesis for how they were created to believe they are of 19th century origin? Third: How was it produced? This is the question you insist I answer first. It is actually the last question, and I'm largely agnostic about the specific mechanism. I don't know, and I don't need to know. My inability to name the specific mechanism doesn't make accurate-translation-of-ancient-plates the more probable hypothesis. If Penn & Teller's audiences can't explain how the trick works, that isn’t evidence that it was done with real magic.

My specific claim wasn’t that there was anything in JS’s specific background that screams he was capable of writing this. My claim is that if you look at the book, you can tell that it was written by an 18th or 19th Century English-speaking American protestant. Hold on. I thought the hypothesis was that it was written by a Mesoamerican. Now you are claiming it was written by somebody from the ancient Near East? Sure, Mormon was allegedly a descendent of people from the ancient Near East. But that was a thousand years earlier. If you trace my ancestry back a thousand years, you’d see I’m a descendent of Celts, Anglos, Saxons, Germans, and Scandinavian Vikings. But like Mormon, those are people from different cultures, from a thousand years earlier, on the opposite side of the world. If you were trying to figure out if I’m an authentic 21st century American, why would you look for clues of me being an 11th century Britain? Let’s get real here. Joseph Smith claimed that the Bible should only be believed to the extent it was translated correctly, but that the Book of Mormon can be believed without that qualification. Wasn’t the point of the Isaiah chapters being included to show how they were translated better than the KJV? Joseph Smith was right that there were translation (which in this context includes errors introduced by copying old manuscripts). But isn’t the issue now settled that whoever wrote the BoM not only failed to fix the errors that were in the KJV, but he also broke things that were translated just fine? And isn’t it clear that whoever wrote it was relying on the KJV with particular distrust of the italicized words. Doesn’t that scream something to you about when it was written? Here’s an obvious example. After the Roman Empire fell, a monk by the name of Dionysius Exiguus lived in the city of Rome. Dionysius was trying to calculate when Easter was when he had a radical idea. Wouldn’t it be helpful if a calendar was created that started with the first year as the year Jesus was born? He did the calculations, and decided to name that year Anno Domini 525. Over the next 500 years, this way of numbering solar years gradually spread throughout the Latin Church, and over the next 500 years after that, the idea spread to the Eastern church as well. That is how such ideas spread in the real world. Meanwhile, in Mesoamerica they had a calendar based on cycles with 260 days, 20 days, and such. This Mayan calendar spread across the region, was widely adopted, and was used for 2000 years. You see where I’m going with this. The Nephites not only used the old-world 7-day week, lunar month, and solar year without influencing or being influenced by the local calendars, they also adopted the Anno Domini calendar literally on the day Jesus was born, a full 525 years before Dionysius invented it, and a full thousand years before it was widely adapted in the old world. Doesn’t that scream something to you? More fundamentally, real religions evolve over time. We could talk about how Judaism evolved or Christianity evolved as examples. But a clear one is how Mormonism has evolved over just 200 years. It started out as something very similar to Protestant Christianity, then had Kirtland-style temple worship, Nauvoo-style temple worship, Brigham Young-style temple worship, and all of the variations in temple covenants and styles since then. The second anointing and calling and election made sure were added. Polygamy was added and then removed. Blacks not being eligible for the priesthood was added and removed. The need of gathering the Saints was important and then not. Other than several spinoff groups, nobody thinks about returning to suburban Kansas City to fulfill the prophesy that the Church’s HQ is supposed to be there. The word of wisdom was revealed not as commandment and then became a commandment. Lorenzo Snow popularized his couplet, and then the Church decided is was just a couplet and abandoned the first half of it as something we don’t know much about. My point of all that is that real religion evolves. But in the Book of Mormon, the religion doesn’t evolve. Not only is the religion static for 1,000 years, the static religion is 19th-century American Christianity. Doesn’t that scream something to you? Would you like to talk about the Book of Ether? Joseph Smith (or whoever wrote it in the 19th century) was steeped in Bible culture, and when he wrote a book that deliberately tried to sound like the Bible, to some extent he succeeded and included things like “poetic parallelism” and “the prophetic call formula.” That’s all great. But it isn’t evidence that the Book of Mormon was written in the 5th century in America, and it whatever it’s evidence for, it’s evidentiary force is orders of magnitude less powerful than what I described above. The Book of Mormon is full of specific prophesies about Columbus, the American Revolution, and a choice seer named Joseph, son of Joseph. But after that, the prophesy becomes vague and disappears--it doesn’t prophesy that a prophet named Spencer would extend the blessings of the priesthood and temple to everybody, regardless of race. Doesn’t that scream something to you? I disagree.

I used that vague phrase deliberately, hoping that it would bias the GPT against my own position. I thought it did a pretty good job of defining what that meant. I included the following phrase in the prompt to guard against the bias you are describing here. Treat arguments for authenticity with the same seriousness as arguments against it. Do not assume at the outset that naturalistic explanations are inherently more probable. We are practicing methodological empiricism, not methodological naturalism. I'm doing so by making very detailed prompts that are intended to constrain how it approaches the question. If I were having a conversational chat with it, I'd expect it to use some latitude to please me. But if I really lock down what I'm asking it to do, I'm moderately hopeful that it will answer the question as asked, using the constraints I give it. You can also go into what ChatGPT calls “temporary chat” which will generate a response in a silo and not archive the solution. Frustratingly, this doesn’t work with the “pro” mode I’ve been using. I guess the idea is that with pro-mode, it’s intended to generate research-grade answers that will be based upon research-grade prompts and not biases from past chats. No problem. I like meta conversations about AI. It's easy to create a ton of low-value words with GPT if you use it that way. I'm hoping that if somebody happens to be interested in what I think, they'll be able to see that in the prompts I ask.

Some AI’s are much better at thinking than others, but in general you are right--AI’s will answer the question given, and if our questions are biased, biased responses will come out. In this case, I was NOT asking for it to explain “why the Book of Mormon is NOT miraculous.” Rather, I was basically asking, “if we assume it is not miraculous, how in the heck was it written?” It’s the difference between arguing whether or not David Copperfield has genuine supernatural abilities, and assuming he does not and asking him how he performed an illusion. When trying to reconstruct (from a naturalistic perspective) how the Book of Mormon came to be, the “sealed portion” seems significant. It’s worth pointing out that I tried to lead the AI to say that it was written by a group and that Joseph Smith was secretly reading a manuscript, but the AI rejected that idea. That indicates it was doing more than merely being a sycophant. Here is my best effort at a neutral prompt. Prompt **Question:** Is the Book of Mormon true? Your task is to evaluate the totality of the *secular* evidence, both for and against this proposition, and produce a single overall likelihood ratio. I will combine your likelihood ratio with my own Bayesian prior (based on personal spiritual evidence) to compute my posterior odds. For clarity: * If my prior probability is 99%, then my prior odds are 99:1. * I will multiply those prior odds by your likelihood ratio. * Your role is to supply the likelihood ratio based solely on secular evidence. Please proceed as follows: --- ### 1. Define the Claim Clearly Define precisely what it means for “the Book of Mormon is true.” Frame it as a binary proposition: either it is true or it is not. Avoid vague middle positions. Make the definition clear enough that the evidentiary question is well-posed. --- ### 2. Compile the Secular Evidence Survey the relevant non-spiritual evidence, both positive and negative. This includes (but is not limited to): * The official and unofficial witness statements * The circumstances and time frame of its production * Linguistic features (e.g., Hebraisms, chiasmus) * Internal coherence and chronology * Historical and archaeological considerations * Literary complexity * Any evidence suggesting 19th-century origin Treat arguments for authenticity with the same seriousness as arguments against it. Do not assume at the outset that naturalistic explanations are inherently more probable. We are practicing methodological empiricism, not methodological naturalism. --- ### 3. Organize the Evidence into Independent Chunks Group the evidence into reasonably independent categories (“chunks”). For each chunk: * Estimate a likelihood ratio * Clearly state what is being compared (P(E | True) vs. P(E | Not True)) * Explain your reasoning Be cautious about independence. If two pieces of evidence are likely correlated, adjust accordingly. When in doubt, err on the side of applying stronger covariance adjustments rather than overstating independence. --- ### 4. Produce a Clear Report (≈1,000 words) Structure the report as follows: **A. Methodology** Briefly explain the Bayesian framework being used, including odds, likelihood ratios, independence, and covariance adjustments. Assume an intelligent lay audience. **B. Evidence Summary Table** List the evidence chunks, the likelihood ratio assigned to each, and how they were aggregated (including any covariance adjustments). **C. Final Likelihood Ratio** Provide the overall likelihood ratio based on the secular evidence. **D. Instructions for Use** Briefly explain how I should combine your likelihood ratio with my prior odds. --- Keep the tone neutral, serious, and analytically rigorous. The goal is not advocacy, but a disciplined evaluation of the secular evidence surrounding the claim. ChatGPT 5.2 Pro Thought for 23m 13s > Here’s a friendlier, message-board-friendly rewrite in plain text (no fancy formulas, no subscripts/summations, no stacked fractions). It keeps the same structure and numbers as my prior answer. A SECULAR (NON-SPIRITUAL) BAYESIAN-STYLE LOOK AT “IS THE BOOK OF MORMON TRUE?” (With a single overall likelihood ratio at the end) WHAT “TRUE” MEANS (BINARY CLAIM) To make this an evidence question, we need a clean yes/no proposition. “TRUE” means: The Book of Mormon is an authentic ancient record describing real ancient peoples and events (roughly 600 BCE to 421 CE), originally written on actual plates in an ancient language (“reformed Egyptian”), and Joseph Smith produced the English text by translating that ancient record through divine means. In other words: the book’s core historical claims are broadly grounded in ancient reality. “NOT TRUE” means: The Book of Mormon is not an ancient record translated from ancient plates. It is a modern (19th-century) production (whether deliberate, subconscious, collaborative, or another natural pathway), and the ancient-American narrative is not a real ancient history. Those are the two bins. No “middle” positions for purposes of this likelihood-ratio exercise. THE BAYESIAN IDEA IN PLAIN ENGLISH A likelihood ratio (LR) answers this question: “How much more likely is it that we would see this evidence if the Book of Mormon claim is true, compared with if it is not true?” LR greater than 1 favors “TRUE.” LR less than 1 favors “NOT TRUE.” LR equal to 1 is neutral. If you have multiple evidence “chunks” that are reasonably independent, you can multiply their LRs to get an overall LR. But in real historical questions, evidence overlaps. Example: “KJV dependence” and “19th-century environment” are related; they shouldn’t be treated as two fully independent punches. So I applied covariance discounts: basically giving some categories partial weight so we don’t overcount correlated evidence. Important: this is not pretending the numbers are objective facts. They are disciplined estimates meant to force clarity about what moves the needle and by how much. EVIDENCE CHUNKS (RELATIVELY INDEPENDENT) AND MY ESTIMATED LRs I grouped the secular evidence into six chunks. For each chunk I give: what it includes, a rough LR before discounting (how it points), whether it got discounted for overlap, the “effective LR” used in the final multiplication. Chunk 1: Witness testimony and plates handling Includes: Three and Eight Witnesses statements; persistence and nuance of their testimony over time. Reasoning in plain terms: If the record is true, you might expect witnesses. If it is not true, witnesses are also plausible via religious visionary experience, group reinforcement, social pressure, etc., especially given the tight-knit networks involved. So this is real evidence, but not uniquely diagnostic. Assigned LR: 3.0 (about 3-to-1 in favor of TRUE) Discount: none Effective LR used: 3.00 Chunk 2: Production circumstances Includes: Rapid dictation in April to June 1829; scribes; reported translation method; manuscript features consistent with dictation. Reasoning: If TRUE, rapid dictation via an unusual method is not surprising. If NOT TRUE, rapid oral production of a long structured text is harder but not impossible (planning, improvisation skill, biblical language scaffolding, prior mental composition, etc.). So this modestly favors TRUE, but overlaps somewhat with witness/setting issues. Assigned LR: 2.0 (2-to-1 in favor of TRUE) Discount for overlap: yes Effective LR used: 1.74 Chunk 3: Literary and linguistic “ancient-like” features Includes: chiasmus/parallelism claims; “Hebraisms”; complexity and internal coherence. Reasoning: If TRUE, you might expect Semitic-like structures. If NOT TRUE, biblical imitation plus human pattern-finding in long texts can also generate these results. Methodology matters a lot (selection effects, what counts as a “hit,” etc.). So I treat this as slight evidence for TRUE, and I discount it because it’s not independent of the biblical-style issue. Assigned LR: 1.3 (slight support for TRUE) Discount: yes Effective LR used: 1.17 Chunk 4: King James Bible dependence and biblical-critical tensions Includes: heavy KJV phrasing; KJV-specific quotation patterns; Isaiah compositional dating issues (for example, the mainstream scholarly view that portions of Isaiah are later than the timeframe the Book of Mormon narrative would imply); New Testament-like phrasing. Reasoning: If NOT TRUE, KJV dependence is strongly expected (it’s the English Bible of Joseph Smith’s environment). If TRUE, believers can argue for a “translation into familiar scriptural English,” which reduces the surprise. Still, the detailed pattern (including KJV-specific forms and the Isaiah critical-dating tensions) looks substantially more expected under NOT TRUE than under TRUE, on secular assumptions. Assigned LR: 0.05 (about 20-to-1 against TRUE) Discount: yes (overlaps with chunk 6) Effective LR used: 0.106 Chunk 5: External fit to ancient Americas (archaeology, material culture, genetics) Includes: broad issues like animals (horses), metallurgy (iron/steel), crops (wheat), writing systems, population genetics, and the general absence of a clear archaeological “match” at the scale many readers naturally infer. Reasoning: If TRUE, one can propose limited-geography models, translation “loan-shifts” (calling something “horse”), and demographic swamping that hides small Near Eastern ancestry. Those moves keep the probability from being literally zero. But taken together, the overall external fit looks much more like what you’d expect if the narrative is not describing an actual ancient Israelite-derived civilization in the Americas at the described scale. Assigned LR: 0.01 (about 100-to-1 against TRUE) Discount: none (this chunk is fairly distinct) Effective LR used: 0.010 Chunk 6: 19th-century milieu and alternative-origin signals Includes: parallels to contemporary texts and ideas; debates over “View of the Hebrews” and “The Late War”; stylometry arguments; and the broader point that there are plausible 19th-century pathways that could generate a book like this without ancient plates. Reasoning: None of these is a single knockout. A lot depends on contested methods, and “parallels” are easy to overstate. Still, the existence of plausible naturalistic pathways and a dense 19th-century cultural ecosystem that “rhymes” with the text pushes somewhat toward NOT TRUE. I discounted this because it overlaps with the KJV/biblical-environment chunk. Assigned LR: 0.30 (about 3.3-to-1 against TRUE) Discount: yes Effective LR used: 0.405 SUMMARY LIST OF THE NUMBERS USED Effective LRs (after discounts where applied): Witness testimony and plates: 3.00 Production circumstances: 1.74 Literary/linguistic features: 1.17 KJV dependence and biblical-critical tensions: 0.106 External fit to ancient Americas: 0.010 19th-century milieu and alternative-origin signals: 0.405 Overall LR is the product of those effective LRs: 3.00 times 1.74 times 1.17 times 0.106 times 0.010 times 0.405 = about 0.00262 FINAL OVERALL LIKELIHOOD RATIO (SECULAR EVIDENCE ONLY) Final secular likelihood ratio: about 0.0026 How to interpret that: An LR of 0.0026 means the secular evidence (as modeled here) is about 1 divided by 0.0026 in favor of NOT TRUE, which is roughly 380-to-1 against TRUE. Put differently: under this particular evidence model, the secular record leans strongly toward a modern-origin explanation, mainly driven by the “external fit” chunk and (secondarily) the KJV/Isaiah-type issues. HOW YOU USE THIS WITH YOUR OWN PRIOR (INCLUDING SPIRITUAL EVIDENCE) Step 1: Convert your prior probability into odds. Odds are just “for” versus “against.” Example: 99% prior means 99 to 1 odds in favor. Step 2: Multiply your prior odds by the LR (0.0026). Example: 99 to 1, multiplied by 0.0026, gives about 0.257 to 1 posterior odds. Step 3: Convert odds back to probability. If odds are A to B, probability is A divided by (A plus B). Example: 0.257 to 1 becomes 0.257 divided by (0.257 + 1), which is about 0.20 (about 20%). That’s it: prior odds times LR equals posterior odds.

I'm a Bayesian methodological empiricist. My objective is to form opinions in line with the evidence, with credence levels proportional to the strength of the evidence. From my perspective, the Book of Mormon screams that it was written in 19th century America. It isn’t just that there a few issues that give me a hangup. It’s that the totality of the evidence points in one direction in an overwhelming way. It’s not that I’m opposed to believing in the supernatural on philosophical grounds. It’s that I believe the evidence isn’t there.

I don’t know. On the one hand, nobody could have written it. But on the other hand, somebody really did. And in this regard, the Book of Mormon isn’t unique. How could somebody without any training in history, no internet, no computer, and not even a research staff write a 560,000 word book with an enormous scale with over 580 named characters, internal coherence, seamlessly blending fiction, philosophy, and military history, extreme psychological realism, and hundreds of coherent character arcs across all of this. And completing it by hand in only 6 years. Creating such a book is absolutely impossible. Yet Leo Tolstoy did in fact write War and Peace. There are endless examples of people doing extraordinary things. The fact that they are extraordinary doesn’t mean they weren’t done. Saying it was impossible to write the Book of Mormon is like saying it’s impossible for the Seahawks to win the super bowl. Maybe so, but that doesn’t change the fact that it happened.

I submitted the following prompt: Prompt In this analysis, you looked at the two official witness statements, but didn't talk about their subsequent affirmations and discussions about their testimonies. Richard Lloyd Anderson wrote a book about this. What are your thoughts about their subsequent statements? How much evidentiary weight does this cary? Personally, I would have loved for a skilled attorney to grill each of them in a 6-hour deposition and get some clarity on exactly what they say happened. Please limit your response on this to ~300 words. Paraphrase of Answer This is in my own words so-as not to post AI content Anderson makes a valid point that the witnesses reaffirming their testimonies for the rest of their lives adds evidentiary weight. But as to what depends on the specific hypothesis we’re talking about. It’s pretty good evidence that they sincerely believed they had “an extraordinary witness experience.” But the specifics of what that experience really was is a bit harder to nail down, and it isn’t much weight at all that this artifact that they witnessed and that subsequently disappeared is of ancient origin. The AI agreed with me on my “deposition instinct.” Performing a really good deposition is a rare skill, and something like this would really help us determine what, exactly, they are claiming they saw, handled, and hefted. Without that, we are assuming some details which might not be completely accurate.

The thought experiment I do is taking a chapter from the Book of Mormon, translating it to modern English, and then handing it to a Mesoamerican scholar and saying, “I was told this was a translation from an ancient Mesoamerican civilization. What do you think? Is that plausible?” If they said anything in the direction of, “Wow, this fits into the time and place of the Mayans (or whomever) beautifully,” I’d be impressed. Note that according to this test, things like noting that on a map the Arabian coast south-south-east, Mediterranean olive growing details, and Hebraisms are all misses, not hits.

I wasn’t aware that bots were banned. My personal philosophy with AI is that I shouldn't use them to churn out massive amounts of content with little effort. I don’t want to delegate my thoughts to them--if somebody wanted to chat with a bot, they are free to do so. My main contribution with the AI is the detailed prompts that I’m making. I only post the AI’s response because I’m using a high-end AI that not everybody has access to.

My prompt was intended to hint at what I think most likely happened. It’s interesting that the GPT didn’t follow my lead on some of the points. All of the talk about the “sealed portion” makes me think that there really had to have been a physical object that was somehow “sealed”, and that this existed before the writing process began. This is what Joseph Smith had in his possession, was hiding, and what other people glimpsed at in all of the years before April 1829. I don’t know who made it or where it ended up though. Maybe he started making them when he claims to have first met Moroni, and it took him several years to finish? No idea. Likewise, I don’t know where Jimmy Hoffa’s body is and I don’t know how he was killed, but I don’t need to know those things to believe he really was. For me the text is relatively impressive, but certainly not impossible. I find, say, The Stormlight Archive by Brandon Sanderson to be waaayyyyy more sophisticated, complicated, interlocking, distinct, compelling, and long. And Sanderson cranks out these books incredibly fast without the help of a scribe. A librarian once reverently told me that every book is a miracle in its own way, and I tend to agree. The text of the Book of Mormon is impressive in a way, but also obviously modern and fictional. Hard to do? Yes. Impossible for the author to have done? No. Translation process. Nothing was translated. How was it done? I have a personal bias towards the fanciful, and love the idea of Lucy Mack and Hyrum carefully writing the book and then handing Joseph sheets to put in his hat for him to read. Despite me leading the GPT towards this theory, it rejected it for the theory that Joseph Smith simply dictated it. Given that he had been telling stories about Lamanites for years before he dictated the book, this is quite plausible. Witnesses. I believe that the witness statements were written first, and that they all promised that they would always confirm that story. I suspect the eight witnesses saw the fake plates. When I envision it in my mind, I always see the experience happening at night, tin plates that look gold in the firelight. The story of Martin Harris “seeing” the plates apart from the others and having to tell Joseph Smith when he finally saw them in a vision makes it clear to me that this was a religious experience, not a real-world one. With all of it, I see a pious fraud going on. Anything that leads people to believe in Christ and to do good must be from God.

Default skepticism of the supernatural might play a part in it, but I'm not sure if you are seeing the significant evidence against this. The issue isn't just supernatural translation; it's the idea of Christians running around Mesoamerica for a thousand years, writing on metal plates for that entire time, creating a massive sword-wielding civilization for that entire time, writing in reformed Egyptian for that entire time, speaking with hebraisms for that entire time, and then vanishing without any archeological, linguistic, cultural, or genetic trace. Either that, or being left indistinguishable from the Olmecs and Mayan civilizations. In principle Christians could have lived in ancient America without supernaturalism. The problem is that there is no evidence that they did. It seems the goal posts have shifted significantly on what we should be expecting here. One would think the whole point of the Book of Mormon is that it was written to the Lamanites, who are a remnant of the house of Israel, to show unto the remnant of the house of Israel what great things the Lord hath done for their fathers; and that they may know the covenants of the Lord, that they are not cast off forever. Now modern apologetics is saying that the literal "remnant of the house of Israel" are genetically indistinguishable from the people who settled in America about 10,000 years before the time "the Lord confounded the language of the people, when they were building a tower to get to heaven," but aren't mentioned in the Book of Mormon because, of course the land of promise that was hidden from all other nations is going to be covered with gentiles and that this is so expected there is no need to even mention them. When you look at how real-world religions and people adapt and borrow and evolve, there is a ton of evidence that this story isn't real. I realize that your mileage may vary, but to me its obvious. In general, there is a strong preference towards naturalism because that is what the evidence always indicates.

I generally don't feel the need to explain the "miracle" of how the Book of Mormon was produced anymore than @smac97 feels the need to explain the miracle of how John Keats wrote his poetry. But he expressed interest in seeing such an explanation, so I decided to pose the question to ChatGPT 5.2 Pro. Below is the specific prompt I gave it, followed by its response. GPT 5.2 Pro is a very advanced AI that does quite a bit of behind-the-scenes research and validation before responding. It does in fact create a log of the major steps it went through in evaluating and researching the question and generating the answer, and if you are interested in seeing that log, let me know. Prompt Can you help me formulate the most likely hypothesis for how, specifically, the Book of Mormon was written? I’m talking with an apologist who believes I need to answer the question of how an uneducated farm boy would be able to do what he did without divine help (e.g. the time frame, how to account for the witness statements, the complexity and internal coherence). Please look deeply at all of the evidence surrounding this and come up with the most likely explanation. Consider popular explanations by critics, but also look outside the box and consider fresh perspectives that may not have been articulated. While you should look at all of the evidence, here are some points I’d like you to consider that I think are salient. Don’t assume these points are true or relevant just because I list them. Do your own independent evaluation: 1- In both the text of the Book of Mormon and in the accounts of the alleged 19th century artifact, a “sealed portion” of the plates is a big deal. This detail seems really strange—both that it is a thing at all, and that it is talked about so much. From a doctrinal perspective the sealed portion is irrelevant—it was never translated and published, and never will be. So why so much attention to something so irrelevant? 2- If somebody were to create a counterfeit artifact and ran out of time or patience before writing “reformed Egyptian” on every page, would they “seal” some blank pages to hide from witnesses the fact they were blank? 3- Joseph Smith wrote the witness statements and had the others sign it. This seems incredibly weak compared to the eleven people creating independent, contemporaneous accounts of what they saw. It was as if they were being told what the official story was. 4- The Book of Mormon contains text that says that everything that invites and entices people to do good is from God. Is that a rationalization that the Book of Mormon itself is “from God” even if there are some white lies surrounding where it came from? 5- The translation process seems to be quite performative. Was he putting on a show to make the people around him think he had ancient golden plates that he was translating in a miraculous way? 6- As I recall, for some of the Book of Mormon he put his face in a hat, and for other parts he was sitting alone behind a curtain. Why would he do it that way? 7- “Radio Free Mormon” approaches it from the perspective of a magician. He asserts the hat was white, and that if there was a piece of paper in the bottom of it, he could have read the piece of paper. 8- Various accounts talk about JS actually reading something. They assume he was reading the translation that was appearing on the seer stone or the Urim and Thummim. But was it possible that he was reading pages from a manuscript that had already been written? 9- There are a few dozen Bible chapters that are quoted in the BoM. They have some changes. When dictating Bible verses, was JS reading from the bible and riffing changes, reciting them from memory, or reading verses that somebody else had modified and written down? 10- Joseph Smith wasn’t quite as illiterate as believers like to suggest, and it is possible that his mother and brother may have helped write it. 11- How long and complex is it, really? Is quickly writing a long and complex book a sign of divine help? For comparison, compared to the Book of Mormon how long and complex is the Cosmere? How fast does Brandon Sanderson pump out words in this? 12- Lots of things are unique. That doesn’t mean divine intervention. Herman Mellvile uniquely had the talent and resources to write Moby ****. Is it possible that Joseph Smith (possibly with help) uniquely had the talent and resources to write the Book of Mormon? Again, those are ideas that are salient in my mind, but don’t use them without independently validating their truthfulness and significance. Given the totality of the evidence, including statements of people who formally or informally saw the plates and were part of or witnessed the translation process. What is the most likely explanation for how, specifically, the Book of Mormon came to be? Include discussions of what, if anything, the witnesses most likely saw, and how the text was compiled and dictated. Please write your answer in conversational English in a way that would be suitable to paste onto a message board. ChatGPT 5.2 Pro Thought for 21m 14s > If I had to pick one single, concrete “most likely” non-miraculous hypothesis that explains both (a) how the English text got produced on that tight 1829 schedule and (b) why multiple people signed witness statements and later reaffirmed them, it would look like this: Joseph Smith was the primary author and performer. He produced the Book of Mormon as an oral dictation project (not as a “sit down and write a book” project), drawing heavily on the King James Bible and the religious/folkloric ideas already in his head from his environment. He used a seer-stone-in-a-hat procedure as (at minimum) a concentration tool and (very plausibly) as stagecraft that let him control information flow, prevent scrutiny, and keep the people around him emotionally and socially committed. A physical “plates” object (real but not ancient) may have existed as a prop; and the witnesses’ experiences were likely a mix of tactile exposure to a guarded object plus religious vision/expectation experiences—all later stabilized into official language via short, formal testimony documents rather than independent, contemporaneous essays. That’s the thesis. Here’s how it cashes out point-by-point with the evidence we actually have. 1) The “how could an uneducated farm boy do that?” premise is doing a lot of rhetorical work Two things can be true at once: Joseph Smith did not have elite formal schooling. He was not the cartoon version of illiterate. The “uneducated farm boy” framing often confuses lack of university training with incapable of sustained oral composition. Scholarship on his education doesn’t support the extreme caricature. (Dialogue Journal) Also: the best historical evidence suggests he had been practicing the underlying skill (spinning sacred-sounding narrative about ancient peoples + religious interpretation) for years before 1829. Lucy Mack Smith describes him, well before publication, giving the family extended “recitals” about ancient inhabitants—their travel, cities, warfare, and worship—i.e., exactly the kind of material that later shows up as Book of Mormon “worldbuilding.” (Joseph Smith Papers) So the “how could he do it so fast?” question is already partly answered by: he wasn’t starting from zero in April 1829. 2) The dictation timeline is tight… but not superhuman once you frame it correctly The “translation” that became today’s Book of Mormon was dictated in a concentrated burst in 1829, with the main sprint running April to late June 1829. (Joseph Smith Papers) But here’s the key: dictation is fast. If a person speaks even ~90–120 words per minute when “in the zone” (slower than normal conversation), you can generate huge word counts in surprisingly few total hours of speaking. And you don’t have to speak 6–8 hours a day; you can do it in intense blocks. So the “time frame” is impressive, yes—but it’s not in the category of “physically impossible without a miracle,” especially if (a) the person has rehearsed themes for years and (b) the product includes lots of formulaic language and embedded Bible-style phrasing. 3) What the best sources say about the actual production method is: seer stone + hat + dictation The strongest first-hand-style evidence (including Emma’s account, echoed by other witnesses) is that Joseph dictated while looking into a hat containing a stone—often with the scribe sitting nearby. The Church’s own historical material now states this plainly and quotes Emma describing him dictating “hour after hour” with his face in the hat. (The Church of Jesus Christ) That matters because it changes the cognitive problem: He wasn’t “writing” in the normal sense. He was performing an oral composition with a scribe doing the mechanics. It also matters because it creates a practical mechanism for controlling what others could verify. If the “translation” is happening inside a hat, nobody can check plates, nobody can check source text, and the social group can only evaluate Joseph’s confidence + the spiritual vibe. 4) Why the hat sometimes and a curtain other times? When you look at the accounts, the “curtain” tradition is much stronger in secondhand/hostile retellings and in some Martin Harris-related reporting; while Emma’s account explicitly says she was close by with “nothing between us.” (The Church of Jesus Christ) A very normal, non-mystical explanation for mixed procedures is: Different settings (Harmony vs Fayette, different houses, different people present). Different audiences (Emma vs Martin Harris vs visitors). Different goals (privacy from outsiders vs theatrics vs preventing accidental peeks at a prop object). In a “performance + dictation” model, you’d expect variation like this, because the method is as much about managing witnesses as it is about producing text. 5) The “witness statements” are not what apologists often imply they are This is huge, because apologists often treat the witness pages like “11 independent depositions.” But historically: The surviving manuscript context shows the witness texts are short, formal statements included with the book. In the printer’s manuscript, the signatures are copied (not original autographs), which is exactly what you’d expect if a scribe is preparing a clean printer’s copy. (Joseph Smith Papers) The Joseph Smith Papers explicitly says it’s unknown who wrote the Eight Witnesses statement and unknown whether the Eight signed the original statement (even while also noting later claims that originals existed). (Joseph Smith Papers) So your instinct in point (3) is basically right in spirit: these are not “eleven separate diary entries written the same week.” They’re official-language artifacts that got standardized very early. That doesn’t make them meaningless. But it absolutely changes what kind of evidence they are. 6) What the witnesses most likely experienced (in a “most likely” naturalistic model) The Three Witnesses: best interpreted as a visionary religious experience Even in official LDS history writing, you see acknowledgments that the witnesses described the experience in visionary terms—e.g., “spiritual eyes.” (The Church of Jesus Christ) Also, Doctrine & Covenants 17 frames the experience as something obtained “by faith,” and the JS Papers describes it as a vision occurring not many days after the commandment. (Joseph Smith Papers) In a naturalistic model, the most likely mechanisms are familiar from revival culture: Intense expectation + prayer Group reinforcement (“we’re about to see something holy”) Social pressure not to be the one person who “doesn’t get it” Retrospective certainty hardening over time (“I know what I experienced”) That can produce sincere testimony without fraud in the moment. The Eight Witnesses: likely handled a physical object, but under controlled conditions The Eight Witnesses statement reads like a sensory/tactile claim (handled, hefted, turned leaves). The JS Papers notes that it “reads like a legal document” and emphasizes sight/touch. (Joseph Smith Papers) In the model I’m laying out, the simplest explanation is: they were shown something physical—but that does not require it to be ancient, gold, or engraved with an actual language. A plausible option is a manufactured prop: thin metal sheets (tin/brass-like) bound in rings, maybe some “characters” on some pages, kept partially covered, and crucially: a “sealed” block that can’t be separated. Which leads to your sealed-portion intuition. 7) The sealed portion: it isn’t doctrinally irrelevant, and it’s also practically useful First, doctrinally: in LDS thought it’s explicitly framed as future revelation, and multiple LDS sources treat it as a meaningful idea, not a throwaway. (The Church of Jesus Christ) Second, historically: multiple descriptions of the plates reported that a large portion was “sealed”/immovable, sometimes described “as solid… as wood,” and estimates range from about half to two-thirds. (The Church of Jesus Christ) Now the key “outside-the-box but not crazy” point: A sealed portion is an elegant multipurpose solution whether you’re sincere, deceptive, or somewhere in between: It explains why the artifact is thick without requiring you to expose every leaf. It prevents anyone from demanding, “Okay, open to a random page.” It gives you an answer to “why didn’t you translate everything?” It creates an always-available future promise (“more will come later”) that can’t be falsified now. And yes—your specific suggestion is logically plausible: if a prop-maker got tired of engraving, a sealed block is exactly how you’d hide blank/unworked leaves while still claiming a big impressive volume. (That’s not proven, but it’s genuinely coherent as a fraud-resistant design choice.) Could Joseph have been reading a prewritten manuscript (hat-as-reading-booth)? This is the “magician” hypothesis: put a paper in the hat, read it, and dictate. It’s physically possible. But it has two big problems: It doesn’t actually solve the main authorship problem. If the book was prewritten, you still have to explain who wrote it and how that happened. We have direct testimony (Emma) denying that he was reading from a book or manuscript during dictation. (The Church of Jesus Christ) You can always say “she lied” or “she was mistaken,” but once you do that you’ve weakened the whole “eyewitnesses prove X” style of argument in general. So: I’d file “paper in hat” as possible but not necessary, and not the best “most likely” explanation. A more economical model is: oral dictation with the hat functioning as a focus device + a sacred-tech prop. 9) The Bible chapters: the data strongly supports heavy KJV dependence, not an isolated miracle of parallel ancient text This is one of the biggest “nuts and bolts” points in favor of a naturalistic model. Serious textual work on the Book of Mormon’s Isaiah passages finds that the KJV is the base text, with lots of variants—many involving KJV italicized words (which were translator-supplied words in the KJV). (BYU ScholarsArchive) That pattern is much easier to explain if Joseph’s production process involved: recalling KJV language from memory, riffing while a Bible is present, paraphrasing with occasional conscious edits, or blending quotation + improvisation in Bible diction. It’s much harder to explain if the English words are coming from a pure “ancient source independent of 17th-century English translation habits,” unless you adopt a very specific theory like “God chose to reveal the KJV phrasing (including its quirks) as the translation language.” And note: even apologetic treatments now openly discuss “KJV translation errors” appearing in Book of Mormon biblical language—because the dependency is too obvious to ignore. (FAIR) So when your apologist says “how could he do that without divine help,” a strong answer is: because a lot of the hardest-to-produce ‘ancient’ style is borrowed straight from the most famous pseudo-ancient English text in his world: the KJV. 10) “Complexity and internal coherence” is real—but it’s not the kind that forces a miracle A few points that tend to deflate the “this is too complex to be human” claim: Much of the Book of Mormon is sermon material. Sermons are exactly what fluent religious talkers can produce at length without writing. Formulaic repetition (“and it came to pass,” stock phrases, recurring moral patterns) lowers the cognitive burden. The work shows seams consistent with dictation and later stabilization (and punctuation was supplied by the typesetter in the 1830 printing process, which can create the illusion of tighter structure than the raw dictation would have had). (Joseph Smith Papers) “Internal coherence” is not binary. Human-made epics can be coherent enough to feel impressive, while still containing the kinds of inconsistencies you expect from oral production. So: the text can be “impressive” and still be a product of a talented, motivated, high-control charismatic dictating for weeks. 11) Putting it all together: a concrete “most likely” reconstruction Here’s the whole hypothesis as a single narrative you can hand to an apologist: Years of incubation (early 1820s–1827): Joseph develops a repertoire of “ancient America + Israelite origin + prophecy + revival Christianity” narrative material and performs it orally (family reminiscences support this). (Joseph Smith Papers) Artifact/authority creation (1827–1829): He maintains a hidden object he calls “plates,” tightly controlling access. Whether it’s ancient or a prop, he structures the social world so no one can freely inspect it. The idea of a “sealed portion” does heavy lifting both ideologically and practically. (The Church of Jesus Christ) Text production as dictation (1828, then April–June 1829): The Book of Mormon is produced primarily by oral dictation, with scribes doing the writing. The seer-stone-in-hat method is central in multiple accounts and in modern LDS historical treatment. (The Church of Jesus Christ) Bible integration: When the narrative calls for Isaiah/Matthew-style material, Joseph leans on the KJV language (by memory, by proximity, or by conscious adaptation). Textual scholarship strongly supports KJV-as-base. (BYU ScholarsArchive) Witness experiences and stabilization: A small circle is given carefully staged experiences: a visionary “showing” for the Three (faith/vision language is baked in), and a controlled tactile encounter for the Eight. The testimonies are then formalized into short official statements rather than independent essays, and later retellings reinforce commitment. (The Church of Jesus Christ) That package explains: the speed, the method, why the plates were never freely examined, why the witness documents look like standardized statements, why the KJV fingerprints are everywhere, and why the story remained resilient even after schisms (because the social/psychological commitment mechanisms are strong in religious movements). 12) How I’d phrase the “so what’s your alternative?” answer to an apologist Something like: “I don’t think the best evidence forces a supernatural translation. The evidence fits better with Joseph producing the text by oral dictation during April–June 1829 using the seer-stone-in-hat method, drawing heavily on KJV Bible language and contemporary religious ideas. The witness statements are short, standardized documents (not independent essays), and even the Joseph Smith Papers note unknown authorship/signing details for the Eight Witness statement and show copied signatures in the printer’s copy. The Three Witnesses experience is framed as faith-based and is often described in visionary terms. A ‘sealed portion’ works both as a theological motif tied to Isaiah’s sealed book and as a practical way to restrict inspection. Put together, a talented, motivated charismatic dictating a Bible-saturated religious epic with strong performance control explains the time frame, the witnesses, and the text’s kind of coherence without needing divine intervention.” (The Church of Jesus Christ)

In my last prompt, I did try to bias it against my own position (e.g. "the book is full of indicators that is ancient..."). On a pretty deep level, the AI wants to be helpful, and will go along with whatever you give it. I think version 5.2 is much better than, say, 4o was, but it's still really easy to manipulate. What I thought was hopeful is that it seemed to be a bit skeptical about the strength of the evidence I listed, and said, "if I treat your listed items as moderately supportive but not decisive...", which is different than saying, "my objective evaluation is that the items you listed are moderately supportive." When somebody simplistically assumes pieces of evidence are all independent and starts multiplying priors by a giant string of usually big factors (e.g. Bruce Dale, Kyler Rasmussen), you almost immediately create meaningless posterior probabilities. In abstract, Bayesian reasoning is the only logically coherent way of weighing evidence, but trying to use it mathematically in an unbiased way is tough. The examples from Caesar were a bit underwhelming for me. “Presented him with two hundred thousand asses” made me think of the explanation of Nephite money in that passage before Korihor attempted to bribe Alma. Is that an example of the Book of Mormon fitting or not? You wouldn't expect Mormon tediously engraving all of that onto the plates, but you could assume that Mormon had incredibly specific prophetic knowledge of his target audience and therefore knew to include it.

For your consideration: Prompt I have some follow-up questions. First, when analyzed the English text of the Book of Mormon, did you include the testimonies of the 3 and 8 witnesses, or did you only consider the text that was allegedly translated? Second, what did you mean “straightforward translation”? Did you mean the *method* of translation was done in a straightforward, conventional way, or did you mean that the translation was a straightforward, reliable rendering of the original language into English? Third, did you discount the likelihood that the text is a reliable rendering of an authentic ancient manuscript because part of the Book (the testimony of the 3 witnesses), say it was translated “by the gift and power of God”? Fourth, the book is full of indicators that is ancient. For example, it has all sorts of Hebraisms. Chiasmus. Ancient Jews writing in Egyptian. The NHM altar and its geospatial relationship with Khor Kharfot. Cement. The "seal of Mulek." Barley. Consistent internal chronology and geography. Alma as a male given name. The Allegory of the Olive Tree. Did you consider these ancient elements in your analysis? Would you like to update your probabilities given these facts? Fifth, when we say the book was translated into “English”, we mean it was translated into King James English as imperfectly understood by an early 19th-century American. Did you weigh this being the target language of the translation against the probability of it being a faithful rendering of an authentic ancient manuscript? The things you said we’d expect from a genuine translation are “consistent “translationese,” odd but stable syntax, culturally grounded idioms, and only occasional overlap with early-modern English Bible phrasing unless the translator is deliberately imitating it.” Let’s stipulate that the translator was deliberately imitating early-modern English Bible phrasing and not hold that against the Book’s authenticity. On the other points, can you give us some examples? As a reference source, I propose *Commentarii de Bello Gallico* by Julius Ceasar. Like Mormon, Caesar was a military and political leader, and both Mormon and Caesar lived roughly around the same time. Using examples drawn from William Duncan’s 1750 translation, can you show us several examples of things we expect from genuine translations that are allegedly absent from the Book of Mormon? Please limit the response to about 1,000 words. ChatGPT 5.2 Pro Thought for 20m 59s > 1) Did I include the 3 and 8 witnesses’ testimonies? In my “English-text-only” assessment, I treated the witness testimonies as paratext—signed claims appended to the volume—not as part of the allegedly translated narrative itself. So my linguistic/translation-style judgments were based on the main body that purports to be the translation. That said, if you broaden “based only on the English text” to include everything printed between the covers, then yes: the witness statements become additional claims in evidence (“plates existed,” “we saw/hefted them,” etc.). They can raise (or lower) your overall belief in plates/production, but they don’t do much to answer the narrower question: “Does the English read like a translation of an ancient record rather than an English composition?” 2) What did I mean by “straightforward translation”? I meant “straightforward” in the product sense: a relatively direct, reliable rendering where the English regularly shows telltale fingerprints of coming from a non-English source—stable “translationese,” systematic handling of foreign terms, and routine little clarifications that naturally arise when bridging cultures. I did not mean the method had to be conventional (paper dictionary, known language, etc.). A translation could be produced via an unusual method and still read like a translation—or it could be produced via a miracle and still read like a free theological paraphrase or English composition in archaic style. “Method” and “resulting translation signals” are separable. 3) Did I discount it because the witnesses say it was translated “by the gift and power of God”? No. A claim about how it was produced doesn’t, by itself, count against authenticity in a text-internal translation-style analysis. Where it matters is indirect: if the translator is explicitly aiming at a particular register (e.g., King James–style English), that choice can mask or overwrite some normal translation signals. That’s a real consideration—but it’s about what the target-language choice predicts the English will look like, not about penalizing the phrase “gift and power of God.” 4) Did I consider the “ancient indicators” you listed, and would I update? A few distinctions matter: A. Internal-to-the-English-text features (Hebraisms, chiasmus, Olive Tree allegory, etc.). Yes, those belong in a text-only discussion. But their diagnostic weight is usually modest, because: Hebraisms can be imported simply by writing in biblicizing English (especially KJV-like English), since the KJV itself preserves lots of Hebraic constructions. Chiasmus is real as a patterning tool, but how probative it is depends on how strict your detection criteria are. There’s apologetic literature arguing it’s meaningful, and also serious critique arguing that flexible rules can “find” extended chiasms too easily. Olive tree allegory overlaps strongly with biblical material (notably Romans 11), which reduces how uniquely “ancient American” it looks as evidence. B. External correspondences (NHM/Nahom, Khor Kharfot, barley, cement, “seal of Mulek”). Strictly speaking, these are not “based only on the English text,” because they require archaeology/epigraphy and geographic proposals outside the text. But if we’re now allowing them as additional evidence, they can move probabilities—if they’re solid matches rather than ambiguous correlations. Very briefly, why I’d still weight several of these cautiously: NHM/Nahom: the altars show consonants (NHM) in a South Arabian context; the match to “Nahom” depends on vowel reconstruction and on whether it’s a place name vs. tribal name, etc. Proponents and critics explicitly debate those points. “Seal of Mulek”: the argument depends on reading an unprovenanced stamp seal as “Malkiyahu, son of the king” and then identifying that with “Mulek.” Even the pro presentation frames it as arguable/conceivable rather than a slam dunk. And more generally, unprovenanced seals/bullae are a domain where authenticity disputes and forgery concerns are a known issue. Barley: it’s true that “barley-like” native Hordeum exists in the Americas, and archaeobotany literature discusses domesticated “little barley” (Hordeum pusillum) in the U.S. Southwest. But mapping the English word “barley” in the Book of Mormon onto specific New World taxa and Book of Mormon geographies is where the evidential strength tends to get murkier. Would I update? If I (i) accept your stipulation that KJV imitation is intentional and (ii) treat your listed items as moderately supportive but not decisive, I’d update upward some—but not to “likely.” Concretely: I’d move my odds from “extremely unlikely” to “still unlikely,” roughly in the neighborhood of well under 1% for (1), and lower still for (2), because (2) adds the extra burden of the narrative being straightforward history of real peoples over a millennium (which is a much stronger claim than “some ancient core existed”). 5) Does “KJV English as imperfectly understood by an early 19th-century American” help the case? It helps explain why the target text would sound biblically archaic without requiring that the English itself be ancient. But there’s a tradeoff: the more you lean into “the translation is intentionally filtered into familiar KJV-style idiom,” the more you’re describing something closer to a dynamic/revelatory rendering than a philologically transparent translation. That doesn’t make it false—but it reduces how much we should expect the English to preserve routine source-language fingerprints. Examples from William Duncan’s Caesar: common translation signals Using Duncan’s Caesar (as you suggested), here are several “this is what translations often look like” features—illustrated in the English itself: A) Frequent “bridging” glosses between cultures/languages Caesar/Duncan gives the reader little orientation handles like: “called in their own language Celts, in ours Gauls” This kind of “they say X / we say Y” scaffolding is extremely normal when a text is genuinely crossing languages and cultures. Why this matters for the Book of Mormon comparison: even granting KJV style, I’d still expect more systematic bridging glosses for institutions, technology, flora/fauna, and social categories across a thousand years and multiple peoples—unless the translation is intentionally smoothing nearly everything into familiar biblical categories. B) Retention of foreign technical terms, units, and “untranslated residue” Translations commonly leave some culturally specific terms in place (or translate them very literally), producing controlled “foreignness.” Duncan’s Caesar does this constantly: “presented him with two hundred thousand asses” (a Roman coin/unit, not converted into English money) detailed engineering/measures: “pits of three feet deep…,” plus “crows-feet… which the soldiers called spurs” Book of Mormon comparison point: in a “codex on plates” scenario, I’d expect more moments where the English shows either (i) persistent loanwords, (ii) consistent conversions (“their measure is… in our measure…”), or (iii) translator notes/glosses around culture-bound items. C) Ethnographic “sidebars” that feel like an author reporting a foreign society For example, Caesar’s Druids passage reads like observational ethnography: “The Druids… interpret the will of the gods.” A long authentic record (especially with wars, migrations, and diplomacy) tends to contain lots of incidental ethnographic texture of this kind—often not “preachy,” just descriptive. D) Translators often explain their editorial/translation policy Duncan explicitly tells the reader what he’s doing (names retained, geography index, etc.). That’s not “proof” of anything by itself, but it’s characteristic of real translation culture: translators worry about mapping problems and tell you how they handled them.