Jump to content

Wordprint Studies


Jerubaal

Recommended Posts

The current thread started by David Bokovoy, and Uncler Dale's comments in it, made me think back to the copy of "Echoes and Evidences of the Book of Mormon" that I picked up a few days ago, which discusses wordprint studies as an almost perfect way of detecting same or different authors.

http://www.fairwiki.org/index.php/Book_of_...rdprint_studies

According to these studies, the books of the BoM have different authors, and Spaulding, JS, and Rigdon do not match up. This is manly directed at Uncle Dale, but what say ye all?

I've got to go, and won't be able to comment for a few hours.

Link to comment

The current thread started by David Bokovoy, and Uncler Dale's comments in it, made me think back to the copy of "Echoes and Evidences of the Book of Mormon" that I picked up a few days ago, which discusses wordprint studies as an almost perfect way of detecting same or different authors.

http://www.fairwiki.org/index.php/Book_of_...rdprint_studies

According to these studies, the books of the BoM have different authors, and Spaulding, JS, and Rigdon do not match up. This is manly directed at Uncle Dale, but what say ye all?

I've got to go, and won't be able to comment for a few hours.

My short reply is that none of the individula books of the BoM will match up conclusively with the known

writings of Smith, Spalding, Rigdon, Cowdery, Pratt, etc. ----- because each BoM book is a composite text.

We have to get down to a more detailed level of inspection here. Just like, were we to go looking for Davidic

authorship in the Book of Psalms, we would have to look at each psalm, and at the individual paragraphs

within each psalm.

My own studies show me that the RLDS BoM segment beginning with Alma XX and ending part way through

Helaman II, is of Spalding authorship -- as are a few shorter segments, in other parts of the BoM.

Were the computerized word-printers to go back and re-check their findings, based upon a more detailed

inspection of the BoM text, I'm sure that their results would agree with my own.

And, in fact, in a recent (but not yet published) word-print anaylsis, conducted at a major US University,

the "non-contextual word-markers" for Spalding, Rigdon and Cowdery have been charted out in the BoM

with practically no overlap. That is to say, certain parts of the book clearly match the known word useage

of one of these writers, and others parts clearly match the wording of one of the other two writers. Where

one of the three has high "positive" results, the same computer analysis shows the other two writers to

have high "negative" results.

The latter third of Alma and the first 1/4 of Helaman show up, in this new word-print study, as matching

rather closely my own findings. That is to say, Solomon Spalding wrote the textual block beginning with

Alma XX and ending part way into Helaman II, and neither Sidney Rigdon nor Oliver Cowdery added any

substantial input for that block (other than one linking paragraph, which charts out as a Cowdery passage).

Should any of this affect the faith of the Latter Day Saints?

Not at all -- if your faith is in Jesus' Gospel, in continuing revelation, and in the covenant relationship

of the Saints to The Almighty, you can pass over this sort of scholarly stuff with your testimony intact.

Uncle Dale

samples here:

http://solomonspalding.com/SRP/AlmaSamp.htm

http://solomonspalding.com/SRP/Alma20c.htm

Link to comment

I'll be waiting for the study to be published, and we'll see if it fits what you say.

Hopefully it will -- though I doubt it will make any mention of me. And I doubt it will voice its consclusions

in my own style of rhetoric.

Probably the LDS scholars will simply yawn, and say that it is but one more example of controversial

Gentile opinions, in the context of a topic that has long been surrounded with controversy.

But, in the end, it is not the Mormons who will be the most influencable audience for such reports --

it will rather be the non-Mormon writers of reference books and contemporary articles (who nowadays

seem to largely accept the notion that Smith himself was the author of the entire text).

UD

Link to comment

Uncle Dale:

And, in fact, in a recent (but not yet published) word-print anaylsis, conducted at a major US University, the "non-contextual word-markers" for Spalding, Rigdon and Cowdery have been charted out in the BoM with practically no overlap.

Any hint as to which anti-mormon university that might be?

Link to comment

>My own studies show me that the RLDS BoM segment beginning with Alma XX and ending part way through

Helaman II, is of Spalding authorship -- as are a few shorter segments, in other parts of the BoM.

I seem to remember our discussion on your methodology. I pointed out the flaws in your analysis and you did not disagree. (see my points below)

>Were the computerized word-printers to go back and re-check their findings, based upon a more detailed

inspection of the BoM text, I'm sure that their results would agree with my own.

Uh, Dale, that is exactly what they did, even in the earlier study. They examined specific sections with an identifiable author.

Dale, tell us what validation studies you made on your methodology, and the results. What experts did you consult and colloborate with in formulating your study?

Thanks mucho.

Link to comment

Statistical studies done at BYU in 1979 by Larsen, Rencher and Layton purported to established 24 authors of the Book of Mormon based upon their "wordprints". See the The New Era November 1979. In addition there was something commonly called "the Berkley Study" by late John L Hilton that again used "wordprints" but tried to accommodate the failings of the original study. Both serve as interesting apologetics but fall short of objective scientific analysis.

A different conclusion was reached in A Multivariate Technique for Authorship Attribution and its Application to the Analysis of Mormon Scripture and Related Texts by David Holmes published in Literary and Linguistic Computing by Oxford University Press. Also A Stylometric Analysis of Mormon Scripture and Related Texts Journal of the Royal Statistical Society. Vol. 155, No. 1 (1992), Here are a couple abstracts from it.

An important discovery is the fact that the samples of writings from the various prophets who purportedly wrote the Book of Mormon do not form prophet-by-prophets clusters. The dendrogram in Figure 2 shows that only the two samples from Alma display internal homogeneity... There appears to be no real difference between Alma's vocabulary richness and Mormon's vocabulary richness within the Book of Alma, a conclusion in direct contradiction to the findings of Larsen and the Brigham Young University team. . . .

It is my conclusion, from the results of this research and the supporting historical evidence, that the Book of Mormon sprang from the 'prophet voice' of Joseph Smith himself, as did his revelations and the text of the Book of Abraham. We have seen that the style of his 'prophetic voice' as evidenced by the main cluster of the textual samples studied, differs from the style of his personal writings or dictations of a personal nature.

He is quite respected from his work on the authorship of The Federalist papers and the attribution of De Doctrina Christiana to Milton. Holmes is one of the most prominant literary and linguistic computing scientist worldwide. His work is the best by far the subject.

Phaedrus

Link to comment

BYU statistican G. Bruce Schaalje did write a "Letter to the Editor" criticising Holmes statistical work on the Book of Mormon -here. Dr. Peterson has said in the past that "Dr. Schaalje appears to have won that debate rather decisively."

I highly encourage those interested in this subject to read both studies from Holmes. Both are in respected peer reviewed academic journals. Schaalje's responses can be seen in the same Journal of the Royal Statistical Society linked above.

In looking at the issue I came to the opposite conclusion as Dr. Peterson.

Phaedrus

Link to comment

Uh, Dale, that is exactly what they did, even in the earlier study. They examined specific sections with an identifiable author.

One of which was Jehovah, as I recall -- Yes, there was an attempt to differentiate the authorship by

textual blocks; though the researchers evidently did not take into consideration either the claims for

Nephite redaction nor for Smith's own phraseology coloring the text. It was a first step -- but I was

not epecially impressed; nor have scholars in general pointed to this differentiation as anything reliable.

In fact, the subsequent article in Sunstone, along with the results of the study conducted in England,

indicate quite different conclusions among the "experts."

'Wordprints' Reexamined," Sunstone, 6 - March-April 1981 - 15-21

Dale, tell us what validation studies you made on your methodology, and the results. What experts did you consult and colloborate with in formulating your study?

Thanks mucho.

My original work was carried out "by hand" in 1979-80, and reported in 1980-82, to totally uninterested

RLDS and LDS audiences. Dr. Lester Bush, Jr. advised that I not publish my work -- so I did not.

In fact, I also took his advise and went on an overseas mission -- and never returned to the US mainland.

I donated all of my research materials to the RLDS Archives, LDS Archives, Dartmouth College and the

University of Utah, and never publicized any of my research until the last few years (and then only in a

very limited sort of way, on the web).

My original work in BoM source critical analysis was only one small part of my graduate studies -- and not

a part that I ever actively promoted and publicized among the "experts." I think you've already read the

more pertinent papers:

http://solomonspalding.com/SRP/SRPpap10.htm

http://solomonspalding.com/SRP/SRPpap11.htm

http://solomonspalding.com/SRP/SCIOTA/SRPpap12.htm

http://www.mormonstudies.com/part1.htm

http://solomonspalding.com/SRP/MEDIA/SRPpap16.htm

http://solomonspalding.com/SRP/ColorSMS.htm

One graphic summary of my findings is depicted in the color-coded chart, I've shown here before:

BOMcolr1.jpg

In the chart, the BoM textual block extending from Alma XX to Helaman I (in the 1830 edition) is shown

in red, with a few very short internal breaks. THis is the part of the BoM which I have before called the

most "Spaldingish" text (a term B. H. Roberts first coined -- not me).

You ask what "validation studies" have been carried out, regarding my findings, by competent experts.

As I said previously, a study which is corroborative (though not carried out as a "validation") has been

submitted for professional publication and should appear in a computer statistical journal next year.

I have been encouraged, by the pending appearance of these independent study results, to go back

through the 1830 edition of the BoM and document my earlier findings, on a page-by-page basis. I'm

currently doing that with the Alma XX to Helman I block of "Spaldingish" text, and am finding the word

overlap to run between 92% and 96% per BoM page:

http://solomonspalding.com/SRP/Alma20c.htm

I can also provide here a very short indication of how my word overlap charting corresponds with the

charting of non-contextual word-print analysis for Solomon Spalding in that interesting textual block:

StatAlm5.gif

All you see here is Alma XX and XXI from the 1830 edition -- with preliminary word-print charting shown

in red, and with word overlap percentages, by page, indicated in the black graph line, near the top. I will

share more of this recent charting, once I get the images posted on my web-site:

UD

Link to comment

Folks,

The Holmes study was addressed in G. Bruce Schaalje, John L. Hilton and John B. Archer, "Comparative Power of three attribution techniques for differentiating authors," Journal of Book of Mormon Studies 6/1 (1997): 47-63.

http://farms.byu.edu/display.php?table=jbms&id=136

See also John L. Hilton, "On Verifying wordprint studies: Book of Mormon authorship." In Noel B. Reynolds, ed., Book of Mormon Authorship Revisited (FARMS, 1997), 225-53.

Danite

Link to comment

Dale, very very pretty.

But, alas, I see nothing on validation of your methodology. This whole thing is colored (pun intended) by subjectivity -- no validation, no indepent checking -- can your results be replicated?

Submission is not the same thing as published.

Link to comment

Dale, very very pretty.

But, alas, I see nothing on validation of your methodology. This whole thing is colored (pun intended) by subjectivity -- no validation, no indepent checking -- can your results be replicated?

Submission is not the same thing as published.

No problem -- when you have some free time, you can look at the texts, compile a list of the words in

common, and chart out the percentage of word overlap for yourself.

Remember also my purpose -- to convince the RLDS leaders that they were wrong, when they told me

that there was "absolutely no resemblance between Spalding's writings and the RLDS BoM."

I think I have demonstrated my intended purpose. If others wish to move on to other purposes, using

the same sort of research, that is fine with me.

Take your best shot, Charlie.

UD

Link to comment

It is my conclusion, from the results of this research and the supporting historical evidence, that the Book of Mormon sprang from the 'prophet voice' of Joseph Smith himself, as did his revelations and the text of the Book of Abraham. We have seen that the style of his 'prophetic voice' as evidenced by the main cluster of the textual samples studied, differs from the style of his personal writings or dictations of a personal nature.

He is quite respected from his work on the authorship of The Federalist papers and the attribution of De Doctrina Christiana to Milton. Holmes is one of the most prominant literary and linguistic computing scientist worldwide. His work is the best by far the subject.

Phaedrus

Link to comment

Folks,

The Holmes study was addressed in G. Bruce Schaalje, John L. Hilton and John B. Archer, "Comparative Power of three attribution techniques for differentiating authors," Journal of Book of Mormon Studies 6/1 (1997): 47-63.

http://farms.byu.edu/display.php?table=jbms&id=136

See also John L. Hilton, "On Verifying wordprint studies: Book of Mormon authorship." In Noel B. Reynolds, ed., Book of Mormon Authorship Revisited (FARMS, 1997), 225-53.

Danite

Indeed. The Holmes study was shown to be flawed, and, as I recall, Holmes himself ultimately acknowledged it and ended up adopting the more refined methods employed by the later Berkeley study.

Unk:

So, you claim that there has been a wordprint analysis of the BoM that has: a) established that it was written by a single author and cool.gif established that it was written by (presumably) Sidney Rigdon. Fascinating! I'll be anxiously anticipating the formal publication of those study results ...

Link to comment

>chart out the percentage of word overlap

OK, you need to validate that "percentage of word overlap" is a valid measure in determining authorship. Your fundamental assumption is, IMO, flawed.

You need to test this specific methodology with other texts to validate your methodology and fundamental assumptions.

There is an old saying, Garbage in Garbage out.

And I will leave that task to yourself.

Link to comment

Unk:

So, you claim that there has been a wordprint analysis of the BoM that has: a) established that it was written by a single author and cool.gif established that it was written by (presumably) Sidney Rigdon. Fascinating! I'll be anxiously anticipating the formal publication of those study results ...

No, you have inadvertantly conflated two of my past statements. It was the study in England that concluded

there was but a single author. Although conducted by reputable researchers and published in a reputable

journal, the LDS scholars have not accepted those findings. I've heard that some RLDS/CoC leaders have

agreed with the possibility that its conclusions are vaid -- but I've seen nothing published by them.

The second study was an entirely different computerized word-print analysis, which will be published next

year. This second study differentiates the BoM text in terms of how much each chapter does of does not

correspond to the known writings of three men (Spalding, Rigdon, and Cowdery).

Although I welcome this new study, my problem with it, is that the word-printing was conducted on the basis

of modern LDS chapter divisions (which were concocted by Orson Pratt), and that those blocks of text are

not necessarily uniform textual entities. I would have prefered to have seen a page-by-page analysis of

the 1830 BoM text. However, the word-print experts inform me that a single page from that edition of the

BoM is too short for meaningful analytical study.

There are some ways around this problem, however. Portions of the 1830 text's chapters can be examined

under various chapter-division systems, and the study results can be averaged or otherwise joined together

to give us something closer to a page-by-page differentiation. Perhaps that will be done in the future.

My own work on the text has not been the study of non-contextual word-prints, but rather the study of

common vocabulary and common word-string occurrence.

What I find useful (and what was marginally demonstrated in the chart-section I showed Charlie, here) is

that the chapter-by-chapter wordprinting for Spalding, corresponds quite closely with my own page-by-page

word overlap charting.

Charlie did not see that interesting correspondence, and I do not blame him for that --- I could only show

a very small section of the text now. I'll work up some more extensive charts next month.

Uncle Dale

Link to comment

>chart out the percentage of word overlap

OK, you need to validate that "percentage of word overlap" is a valid measure in determining authorship. Your fundamental assumption is, IMO, flawed.

You need to test this specific methodology with other texts to validate your methodology and fundamental assumptions.

There is an old saying, Garbage in Garbage out.

And I will leave that task to yourself.

Again, you are missing my purpose, Bro. Dowis --- My graduate studies were conducted under the

sponsorship of certain RLDS leaders, who encouraged me to go to University and examine the evidence.

When I reported my findings, Elder Wayne Ham and other upper level RLDS "experts" informed me that

my studies were worthless, because "everybody knows that there is absolutely NO similarity between

Spalding's writings and the Book of Mormon."

My task, after that dismissal from high saintly authority, has been merely to demonstrate the actual

level of textual correspondence of the Book of Mormon and Spalding's known writings.

None of that is "garbage," and there is no reason for you to call it that. What you can say, is that you

find the degree of textual similarity too weak to indicate any substantial relationship -- or some such

thing. But I resent your calling my research "garbage."

Uncle Dale

Link to comment

>My task, after that dismissal from high saintly authority, has been merely to demonstrate the actual

level of textual correspondence of the Book of Mormon and Spalding's known writings.

Your intention is irrevelant to whether your study is valid. If you are going to report the results of your study, be prepared to have it critiqued.

If you don't like what I have to say, defend the study with facts, not emotional outbursts.

>But I resent your calling my research "garbage."

GIGO is a computer term meaning that if you input invalid data (garbage) into the computer, regardless of the infinite computing capacity of the computer, it will only produce invalid output (garbage). In this case, if your assumptions are flawed, if your methodology is faulty, your computations will only produce........

A validation study to test methodology is a pre-requisite to drawing any conclusions from the study's data. You seem to have a problem with that concept.

Without validation of the methods, your data is worthless (garbage). Hilton, et al, performed validation studies on their methodology.

I find your study and conclusions insulting. Prove me wrong.

Link to comment

...your data is worthless (garbage... I find your study and conclusions insulting. Prove me wrong.

I'll leave this one for the mods to decide.

If they conclude that I'm insulting to you, then I'll be happy to take a six month vacation from FAIR,

and leave you to enjoy my extended absence.

UD

Link to comment

Uncle Dale, I'd hate to see you leave. I find your study interesting. I think it would be more interesting if your computer-wiz friends would also do a comparison of the Book of Mormon with a Charles ****ens novel or mabye another novel that is like Spaulding's but not written by anyone ever accused of contributing to the Book of Mormon.

Link to comment

Uncle Dale, I'd hate to see you leave. I find your study interesting. I think it would be more interesting if your computer-wiz friends would also do a comparison of the Book of Mormon with a Charles ****ens novel or mabye another novel that is like Spaulding's but not written by anyone ever accused of contributing to the Book of Mormon.

Like Walt Whitman's Leaves of Grass perhaps?

Link to comment

Uncle Dale, I'd hate to see you leave. I find your study interesting. I think it would be more interesting if your computer-wiz friends would also do a comparison of the Book of Mormon with a Charles ****ens novel or mabye another novel that is like Spaulding's but not written by anyone ever accused of contributing to the Book of Mormon.

Like Walt Whitman's Leaves of Grass perhaps?

One portion of their research involved the use of "double-blind" study methods, by which even they

did not know the origin of certain pre-1830 texts that got fed into the computer. I'm not sure if they

will publish any details of that part of their experiment or not.

What was discovered, is that in any lengthy pre-1830 text, that is compared to the known writings of

Rigdon, Spalding and Cowdery, each of those writers shows up as the "highest probability of authorship"

all through the examined text. -- But the degree of difference between the results for the three men is

generally very minor.

In performing the same analysis of the Book of Mormon text, the results come out with very wide margins

of difference. In other words, where Spalding is statistically shown to be the "highest probability of

authorship," Rigdon and Cowdery generally come out as having negative values.

So, to get back to your suggestion -- perhaps rather than just examining a single pre-1830 novel,

the computerized methods should be applied to several hundred e-texts now available -- so that we

can see if the Book of Mormon is unique in this respect, or whether there are other old texts where a

similar pattern of "highest probability of authorship" emerges.

In fact, were I assigned to run some test as part of a critique of these gentlemen's findings, that would be

exactly where I would begin.

Uncle Dale

Link to comment

Archived

This topic is now archived and is closed to further replies.

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...