DiscussMormonism.com

In Episode 9 of his “Estimating the Evidence” series, Kyler Rasmussen thinks his statistical analysis is formidable. He thinks he bends over backwards “for the critics’ sake” by massaging the evidence in their favor, thus coming up with an estimate of the probability of the Book of Mormon being authentic that is “incredibly safe.” Despite working so hard to dampen the evidence, the final number he comes up with for the Book of Mormon being written by a 19th century author is p = 5.24 x 10^-24.

He likens the strength of this argument to heavy dragonplate, and says, "I’m not gonna lie; I have tremendous pity for critics trying to demonstrate the weakness of this particular class of evidence."

The whole argument here is bizarre. The Book of Mormon obviously isn’t really written in Early Modern English. The apologists are pointing to an unfalsifiable theory that the Book of Mormon is written in Early Modern English, but “filtered and managed” to be understandable to a 19th century reader, with enough Bibleish language deliberately added to make the whole thing sound sacred. Kyler says critics are required to provide “a clearly articulated theory that explains the source of this evidence,” while apologists get a pass and can attribute the alleged weirdness to the mysterious ways of God.

On the Interpreter site, I focused my comments on the simple fact that this is a non sequitur. “Alleged evidence of a 19th century book being peppered with words, grammar, imagery, and protestant religious issues from 16th century England simply isn’t evidence that the book is an accurate translation of an authentic ancient Mesoamerican manuscript.”

Here I’m going to distil Kyler’s actual math.

Kyler found eight categories of “archaic and non-archaic syntactic features” that Carmack has curated. These are basically situations where the author has a choice between archaic syntax or modern syntax. For example, a long time ago somebody wrote, “God requireth the law to be kepte of all men.” This use of “of” is archaic. A modern writer would have used the word “by.”

According to Kyler’s numbers, the translator of the Book of Mormon had 3,489 choices of whether to use modern syntax or archaic syntax. The translator chose to use archaic syntax 74.23% of the time.

The same exercise was done for four pseudo-biblical books, and the numbers for these four books were totaled. In aggregate, the pseudo-biblical authors had 713 opportunities to choose between archaic syntax or modern syntax. The pseudo-biblical authors chose the archaic syntax only 5.75% of the time.

Kyler then ran a basic statistical test on this. The test he ran (Pierson’s Chi-Squared Test), evaluates the likelihood that the same statistical distribution produced both sets of results. It turns out that the answer is no—74.23% and 5.75% are far enough apart to know the underlying propensity of the authors to use archaic syntax really is different.

And that, my friends, is the dragonplated evidence that leads critics to despair and increases the odds of the Book of Mormon being true by a factor of trillions of trillions.

There are several problems with this analysis, and I’m going to state an obvious one.

On a fundamental level, Kyler’s math is implicitly asserting with dragonplated certainty that modern books use archaic language precisely 5.75% of the time. The problem is there is no basis for this. Furthermore, his own data proves that this assumption is not true—if you perform a chi-squared test of the propensity of individual books in the pseudo-biblical group that Kyler collectively compares against the Book of Mormon, those books are different from each other, too. For example, Leacock’s Chronicles uses archaic language 11.3% of the time, while Snowden’s Revolution uses archaic language 1.5% of the time. If we plug those numbers into the chi-squared test, we can be 99.999% certain these books are different.

Different authors make different syntactic choices. Chronicles using archaic language ten-times more frequently than Revolution isn’t evidence that Chronicles was really written by a 16th century ghost committee rather than by Leacock. We don’t need miracle or a God to explain this difference. Different authors have different vernaculars. That’s all there is to it.

A similarly constructer statistical analysis would consider the presence of language patterns from two distinct English language centuries, without the sharpshooter assumptions, and show with similar interstellar odds that the Book of Mormon was a fabrication made in the latter of those two centuries.

The analysis is persuasive that the Book of Mormon was not written in a similar way to those other pseudo-Biblical books. That only implies it wasn’t written in the 19th century, however, if you assume that “19th-century pseudo-Biblical writing” is one particular thing with a fixed set of rules that are accurately exemplified by that handful of texts.

Rasmussen simply assumes, and has in no way shown, that any possible 19th-century effort at a Biblical-sounding book would have to be similar in style to his basket of examples. He assumes that if Smith had written such a book, its prose style would necessarily have resembled that of those other books.

Which were written for sale as entertainment rather than as a hoax, by educated professional writers, whose texts were edited for publication.

Gee. Could every single one of those major differences between the other books and the Book of Mormon perhaps all tend to make the Book of Mormon a lot more archaic in style than the others? Would the professional writers have been willing and able to keep their archaism down to an amusing flavor, fairly consistent with the King James Bible? In contrast, could an uneducated fraudster have clumsily ladled on every archaic stylistic quirk he had ever heard in sermons or hymns or Shakespeare or Bunyan, ham-handedly overdoing it wildly as he dictated to a scribe from his hat?

No, instead of that it is astronomically more likely that a native speaker of 16th century English composed the Book of Mormon while employing 19th century vocabulary and concepts. This is Rasmussen’s premise.

Dragonplate? When someone has to oversell their "prowess" perhaps it is a tell as to what that person really thinks about their own point. I'm surprised he didn't use all caps saying "TAKE THAT CRITIC SCUM! HOW COULD JOSEPH HAVE KNOWN!?!"

Sorry Dr. Rasmussen, I think the explanation posited by PG is the better one:

In contrast, could an uneducated fraudster have clumsily ladled on every archaic stylistic quirk he had ever heard in sermons or hymns or Shakespeare or Bunyan, ham-handedly overdoing it wildly as he dictated to a scribe from his hat?

The answer to this question, Dr. Rasmussen, is yes, especially given the utter lack of evidence that there were actual Nephites.

You're a good sport, Billy! Your comments are very well expressed. KR puts a lot of time and effort into being obnoxiously sarcastic and dismissive, but you just roll with it, and stick to the facts.

I noticed Brant Gardner weighed in at the Interpreter comments for episode 9, in support of one of your points:

Brant A. Gardner on September 2, 2021 at 5:32 pm

For the record, I support an ancient Book of Mormon (original). However, Billy is correct that the Early Modern English evidence doesn’t point to antiquity. It points to Early Modern English, and asserts that it was written about that time. The question of whether or not it is a translation still exists. The difference is now when it was translated, not whether.

The “Joseph couldn’t do it” argument simply suggests that someone else did. Skousen has not opined on how it happened, only that he agrees that Joseph wasn’t behind the translation. We have exactly the same problem, but moved a hundred years (or less) earlier.

This points out, once again, the really, really bad statistical analysis KR is doing here. Regardless of what his individual hypotheses are, he always assumes that support of any individual hypothesis, no matter how irrelevant, is equivalent to support of his overarching hypothesis:
"Book of Mormon Authenticity."

That is incredibly bad statistics, to simply combine all these results, and yet he even has a chart where he shows just that. I suppose it is inevitable he found a home at the Interpreter Journal to publish, because no legitimate, peer-reviewed journal would associate themselves with this silliness.

Regarding this topic, here is Kyler Rasmussen quoting and then replying to the OP:

Kyler Rasmussen on September 1, 2021 at 9:08 pm
Hi Billy!

“This episode is so flawed I don’t have the space here to respond fully, and it is so fundamentally unserious I feel little motivation to do so.”

Your exasperation is a pretty good indication for me that I’m doing something right....

Wow. Talk about not being able to read the room. Here is the first point Billy Shears makes:

Billy Shears on September 1, 2021 at 6:02 pm
This episode is so flawed I don’t have the space here to respond fully, and it is so fundamentally unserious I feel little motivation to do so.

A big issue to get out of the way is the consequent possibility of p = 1. Yes, I’m going to cry foul on this one. What this means, specifically, is that before Carmack and Skousen started this project, i.e. before the experiment was conducted, you declared that if the Book of Mormon is true, then there would be exactly as much Early Modern English as Skousen and Carmack found. You then ran the experiment and behold! Your hypothesis about Early Modern English was precisely confirmed!

This is the most dramatic example of the Texas Sharpshooter fallacy I could imagine. The truth is you drew these tiny targets around the holes in the side of the barn after the bullets hit the barn, not before. That is why they are hits.

This is a huge point. Here is KR's definition:

CH = Consequent Probability of the Hypothesis (the likelihood of observing the evidence given a non-Joseph authored Book of Mormon, or p = 1)

Consider what this means. It means that if Joseph Smith did NOT write the Book of Mormon, it is an absolute certainty, with a probability of 1, that it is a type of Early Modern English text. No other options. It can't be a direct hebrew or Reformed Egyptian or mesoamerican translation, or anything else. It MUST be a "filtered and managed" text that is mostly Early Modern English, but written so that 19th century readers understand the Early Modern English. Without exception.

Maybe Rasmussen is punking us. Because the most reasonable, non-supernatural interpretation of this "consequent probability" is that Smith plagiarized the Book of Mormon. The only way he didn't plagiarize is if a supernatural event occurred. He is really painting apologists into a corner, given that this option is not in keeping with current research as Dr. Moore pointed out.

Dr Rasmussen has lost, basically before he started this. And yes, this episode is as bad or worse ( I mean the first vision one?) as any other.

The way he attempts to flip the likelihood completely upside down on the question of Early Modern English presence in the Book of Mormon is just ludicrous.

You are right, Billy, to suggest its flawed and unserious. There's no way to read that and reasonable think, Ok...I mean it may be a stretch but ok. No, the only reasonable response here is to say, that was a complete and utter mess that didn't make sense, nor work logically, from start to finish.

Thanks, Lem!

Going into his arguments more deeply, Kyler is trying to add some mathematical rigor to Stanford Carmack's 2018 paper, Is the Book of Mormon a Pseudo-Archaic Text?

On Carmack's Arguments
In that paper, Carmack compares the Book of Mormon to a handful of pseudo-biblical books and the Bible in terms of how frequently they contain the specific Early Modern English elements that he hunts for in the Book of Mormon. In that study, Carmack concludes:

In a nutshell, the Book of Mormon text exhibits high levels of archaic (morpho)syntax; the pseudo-biblical texts exhibit much lower levels of archaic (morpho)syntax.

A sufficient and accurate knowledge of the form and structure of the earliest text of the Book of Mormon reveals that The Late War pales in comparison with the Book of Mormon in terms of archaic usage. In fact, the other three pseudo-biblical texts are more archaic than Hunt’s text in many different linguistic domains. In view of these linguistic facts, had Joseph created literature like The Late War, or had this pseudo-biblical writing or another comparable text taught Joseph how to fashion older, biblical language (or influenced his dictation to scribes, etc.), the form of [Page 232]the earliest text of the Book of Mormon would be very different. It would be both more biblical and more modern in character, as we find is the case with the four pseudo-biblical writings considered in this study.

I'm not qualified to argue with Carmack about the linguistics, but I do have the following thoughts that may (or may not) be worth considering:

1- The archaic usage he focuses on are based on things he found from his monomaniacal search for such things in the Book of Mormon. This could lead to a bias in his conclusions. For example, are there archaic elements in other books that don't exist in the Book of Mormon, which he is blind to because of their absence in his favorite book?

2- We have no reason to believe an authentic Book of Mormon would have archaic language that wasn't in Joseph Smith's vernacular. That being the case, the fact that it is sprinkled with this archaic language is only a "hit" in the sense that the target was drawn around this amorphous target after the data was collected and analyzed.

3- We have almost no data on how people in rural New England actually talked, especially when they were trying to sound biblical, old fashioned, or religious. The Late War might have less archaic language than the Book of Mormon because Gilbert Hunt was more educated that Joseph Smith.

The hypothesis that remnants of Early Modern English may have survived into sacred verbal speech in New England seems much more plausible than the hypothesis that a ghost committee from the 1600's wrote the Book of Mormon. Ultimately, that is a question for linguists, but with such limited data on how people actually spoke, how can we know for sure?

On Rasmussen's Math
In his paper, Kyler links to some class notes from an undergraduate Sociology class at the University of Utah which gives an example on how to do a chi-squared test. It says:

In the following example, we'll use a chi-square test to determine whether there is a relationship between gender and getting in trouble at school (both nominal variables). Below is the table documenting the raw scores of boys and girls and their respective behavior issues (or lack thereof):

46 Boys got in trouble
71 Boys did not get into trouble
37 Girls got into trouble
83 Girls did not get into trouble

To examine statistically whether boys got in trouble in school more often, we need to frame the question in terms of hypotheses. The null hypothesis is that the two variables are independent (i.e. no relationship or correlation) and the research hypothesis is that the two variables are related. In this case, the specific hypotheses are:

H0: There is no relationship between gender and getting in trouble at school
H1: There is a relationship between gender and getting in trouble at school

The example then goes on to show how you can plug those four numbers into a Chi-Squared test to determine whether or not boys and girls are equally likely to get into trouble.

Simple stuff.

Kyler noticed that you could format Carmack's data into four numbers that correspond to this example. Instead of testing whether boys are just as likely to get into trouble as girls, he tests whether the Book of Mormon is as likely to contain archaic language as a collection of other pseudo-biblical work.

That's great as far as it goes, but he completely misinterpreted what the test means. He thinks critics absolutely, positively, definitely, need the frequency of archaic words in the Book of Mormon to precisely match the frequency of the same words that are in the Bible, or at least in the collection of the four pseudo-biblical works in the study. He describes the result of his test thusly:

The chi-square values in each case are pretty astronomical; χ2(1) = 1186 for pseudo-biblical; 2337 for the King James (note that the values in the parentheses when reporting chi-square values represent the degrees of freedom for the test); that leaves us with probabilities so small that they have negative exponents in the hundreds (p = 6.46 x 10-260 and 5.35 x 10-510). If we lined up all the particles in the universe (of which there are about 1080), treated each particle like a lottery ball, and then had to correctly select the right ball, we would have to pick the correct one at least 3 times in a row to get a similarly improbable outcome. Clearly, we can be confident that the differences in syntactic usage in the Book of Mormon don’t differ from the Bible or the pseudo-biblical works on the basis of chance. It just so happens that the probability for the pseudo-biblical works is somewhat higher (because it has a smaller sample size overall), so for the critics’ sake we’ll ignore the comparison with the King James.

The astronomical chi-squared values don't imply the Book of Mormon is ancient. They imply Rasmussen doesn't understand how to use statistics outside of canned problems in a classroom.

If Kyler's real objective is simply to waste everyone's time, he's doing a wonderful job.

the final number he comes up with for the Book of Mormon being written by a 19th century author is p = 5.24 x 10^-24

If this is accurate then Kyler can also tell us what the probability is that Atlantis was real. The probability of the identity of DB Cooper. The probability for each location of Jimmy Hoffa's body.

Kyler has stumbled onto a literal oracle of endless insight into past history. Or he's just bullshitting and wasting everyone's time.

Lem wrote: ↑

Fri Sep 03, 2021 7:45 pm

Consider what this means. It means that if Joseph Smith did NOT write the Book of Mormon, it is an absolute certainty, with a probability of 1, that it is a type of Early Modern English text. No other options. It can't be a direct hebrew or Reformed Egyptian or mesoamerican translation, or anything else. It MUST be a "filtered and managed" text that is mostly Early Modern English, but written so that 19th century readers understand the Early Modern English. Without exception.

I didn't appreciate this until recently, but this is indeed exactly how the Sharpshooter Fallacy often appears in (bad) Bayesian inference. My example from a few weeks ago here was my claim to be a prophet based on my producing exactly the long string of random characters that was God's specific message (according to my sect's doctrine).

The sharpshooter declares that the probability of the data, given his hypothesis, is one. That's precisely the Sharpshooter fallacy. The accurate analysis, in contrast, has to be accurate about how likely the data really are, given the hypothesis.

Even if the sharpshooter really were an incredibly accurate shooter, that weird random pattern of holes in the barn would have been an unlikely choice of targets for him. Once you recognize this and give his p(holes|sharpshooter) an appropriately low value, Bayesian inference ceases to be a sucker for the Sharpshooter fallacy.

In this case it is clearly very unlikely that exactly this linguistic pattern would occur in the Book of Mormon even if it were revealed by God. Nobody (except perhaps me!*) has suggested a reason why God should reveal the Book in such language.

I still think that the bigger problem with Rasmussen's analysis is that the actual Book of Mormon language really isn't unlikely at all, if Smith clumsily overdid his archaism. But the other side of the equation also needs this big revision, to remove the Sharpshooter Fallacy. In the end, for my money the linguistic analysis of the Book of Mormon only makes it seem more obvious that Smith made it up.

*My Mormon theory, which I have offered on both MDDB and Jeff Lindsay's blog, is that God might deliberately have made the divine translation in a weirdly archaic English dialect, as a divinely perfect way of representing the archaism of the original Nephite Hebrew in its day. The Nephite vernacular evolved in the centuries in the New World, but the faithful scribes who ended up writing the Book of Mormon preserved their old-fashioned Hebrew for religious reasons. God saw fit to recognize their faithfulness in this by rendering their Book into analogously archaic English.

I think this is a darn good theory, myself. But no Mormon Early Modern English enthusiasts seem to have taken it up, perhaps because they fear the geeks even when they bring gifts.