This research extends what I have shown in Textual analysis of Book of Mormon timeline proves it could not have been faked. Therefore, for brevity, I won’t explain here the dataset or the methods used to derive said dataset, because I assume you have already read that essay.
The object of this present essay is to evaluate whether the order of the doctrines presented in the Book of Mormon aligns with the Bible. And to note: I only have done this analysis and the previous analysis on the Book of Mormon timeline because I felt inspired to do so in An Order to the Gospel.
Therefore, here I evaluate whether the order of the gospel – which I interpret here as being represented in part by the principal components produced in the last research – holds constant between the Bible and the Book of Mormon. To do this, as previously described in the last essay, I used AI to produce a list of features for each verse in the Book of Mormon and the Bible, and then extracted the top 20 principal components of these features.
More specifically, the question at hand is whether the data extracted from the Book of Mormon is able to improve the predictive capacity of principal component models on the 264 features extracted for each verse in the Bible, when the models are evaluated on data in the Bible that wasn’t used in training.
Hint – this was the case. But to explain more in depth: when my model is trained, it produces the optimal weights necessary for the top 20 principal components of each verse to have the highest possible predictive power for the 264 features extracted for each verse. When the model is evaluated, it then uses these weights on the features in the validation data to produce the values for the principal components for the validation data, and then it calculates the amount of variance that these principal components explain in the validation data. If the model is good, then the principal components it produces should be able to explain a large amount of the variance in the validation data.
Therefore, if the Book of Mormon has unique data with explanatory power of the structure of the gospel in the Bible, then if I train the model adding data from the Book of Mormon, then I would expect the model to perform better than if it is trained without data from the Book of Mormon.
The specific setup for all of this is: for each sub-book of the Bible, I trained the model on all of the data in the Bible not in that sub-book and then evaluated it on the data within that sub-book – repeating in a loop across all of the sub-books of the Bible. I then did the same thing, but adding to the training data the data from the Book of Mormon.
And what do you know: the data from the Book of Mormon does add explanatory power!
The mean variance of the 264 features for each verse in the Bible – not explained by the model – when the model does not have the Book of Mormon is 166.73. But when the data in the Book of Mormon is included, then this variance goes down to 158.46 – and the p-value that this decrease in variance is significant is very, very low: 1.165 x 10^-91.
And to note: the validation set is only ever from the Bible – so here I show that models predicting the Bible do better when they are trained on both the Bible and the Book of Mormon, than if they are just trained on the Bible.
This means that the data of the Book of Mormon is – on average – able to provide unique insight into each book of the Bible, of which insight is not provided by the rest of the books of the King James Bible.
I don’t think this strictly proves the Book of Mormon in the same way that I think my past research has proven the Book of Mormon is true.
However, if Joseph Smith faked the Book of Mormon, it would have been necessary for him to have an intimate knowledge of the peculiarities of how the gospel is presented in each particular book of the Bible – a knowledge with much deeper insights than could have been found by merely copying the Bible – since if the Book of Mormon was merely a copy of the Bible then it would not actually improve the models. Seeing that Joseph Smith was an uneducated 23-year-old when he translated the Book of Mormon, it seems unlikely that he could have possessed this incredible knowledge, though I suppose strictly speaking this is technically possible, albeit very unlikely.
And to people who say that the Book of Mormon goes against what the Bible says, my reply is: statistically speaking, the Book of Mormon has the same structure of how doctrines are presented as is found in the Bible. Therefore, inasmuch as you are attacking the Book of Mormon, you are also attacking the Bible.
All of the code and data requisite for this result can be found in Textual analysis of Book of Mormon timeline proves it could not have been faked, and the specific file in which I did this calculation is called simply_showign_the_bom_is_true.py.
Leave a Reply