Uncle Dale Posted December 21, 2008 Share Posted December 21, 2008 Again, word-print studies are not like fingerprinting and DNA where "closest" isn't good enough. With word-prints, there is a far greater chance of a false positive (100% chance if the actual author is not included in the study), and no way of determining if it is a false positive.Depends on what me mean by "false positive." If we see one or two chapters allotted a high probability to one of the seven authors tested, and if we suspect a "false positive" result, due to the inherent mechanics of that particular word-print testing method, we can conduct more tests, using somewhat different parameters, data, authors, etc. Such additional analysis should help isolate any "false positive" resulting from a computer error.But there is another sort of false positive -- and that might occur when a certain author's word-print consistently tests out at a very unexpected level. Suppose we could somehow isolate a word-print for Lehi, the father of Nephi, and when we tested that word-print against the BoM text, we saw a 99% probability (out of numerous authors examined) that Father Lehi wrote the Book of Moroni. We would suspect this second sort of "false positive."I think you are blurring the distinction between these two possibilities.So, we aren't talking about a tecnique that can determine authorship to a scientific certainty.We are talking probabilities here -- relative probabilities, at that. The "scientific certainty" that 2+2=4 is rather different from the scientific expectation that an animal that walks and quacks like a duck will, upon close examination, turn out to be truly a duck.Scientific findings based upon probabilities are just that and nothing more. We cannot advertise them as 2+2=4 sorts of "scientific certainty" --- but we can use them to our benefit in conducting further research. At best, it is an educated guess--which is only as good as the sample and control groups included in the studied (which, in this case, is as yet not impressive). It is a matter of faith rather than fact.There are facts involved, to be certain. And there are deductions and conclusions rendered, based upon analysis of the data. There is something of a parallel to "faith" in our accepting any such findings ---- but the scientific method does not rely upon "faith." The next word-print study that comes along may provide rather different results, and some folks may accept those contrary findings as being consistent with their "faith." That is human nature; but it is not Science.I would have more confidence in the study if the "Nephites" were tested.I'd guess, to get to that point, the LDS leaders/scholars will first of all have to produce and communicate to the rest of us, exactly the text samples that they have identified for each of the main "Nephite voices" in the book. This has been attempted in earlier word-print studies. I expect it will come again.I would be even more confident were the original text, rather than the translation, to be tested against other writings of the "Nephites" in their own tongue,Until common consensus knowledge includes verifiable examples of the Nephite dialect of Hebrew, rendered in Reformed Egyptian format, I sincerely doubt such a thing will be possible.and were Rigdon's pre-conversion rather than post-conversion writing to be sampled.If Rigdon wrote the 1824 "Third Epistle of Peter," published by himself and Scott, then THAT text ought to be compared to the Book of Mormon. But we could only be certain it was Rigdon's production, by first of all comparing it to his post-1830 word-print ------ so, that process might harm your confidence in the process as well.Until then, this all makes for amusing intellectualizing, but little more.Depends on whether you are speaking as a Mormon, a "Smith-wrote-it" anti-Mormon, or a disinterested Gentile statistician, out to make a name for himself by either refuting or confirming the Stanford findings. It also completely misses the point of the Book of Mormon and the restored gospel of Christ (i.e. bring us to Christ--which, to me, is where my energies should be focused).Try explaining THAT to some disgruntled Gentile millionaire from San Francisco, who is now willing to shell out a few hundred thousand bucks, to fight the Church on social issues, or Mormon origins, or whatever.But, to each their own.Thanks, -Wade Englund-UD. Link to comment
wenglund Posted December 21, 2008 Share Posted December 21, 2008 Depends on what me mean by "false positive." If we see one or two chapters allotted a high probability to one of the seven authors tested, and if we suspect a "false positive" result, due to the inherent mechanics of that particular word-print testing method, we can conduct more tests, using somewhat different parameters, data, authors, etc. Such additional analysis should help isolate any "false positive" resulting from a computer error.I am not speaking about computer error. I am speaking about sample error. If the sample doesn't include the actual author of the text (and there is no way to know if this has occured or not since this is ultimately what is at issue), then the study will invariably return a false positive (meaning, the wrong person will be attributed, probabalistically, as the author).In other words, using your last analogy, if the testing monitor or the proctor were the actual authors of the answer in the bluebook in question, and were they not included among the 21 students sampled in various permutations of the hypothetical wordprint study, then the study will invariably return a false positive. One of the 21 students would be falsely attributed (based on probability) with the authorship.And, there is no way to know that that isn't the case with the study in question. In fact, given the very limited sample size, drawn in part from texts written after the Book of Mormon was published (and may thus have been influenced by the Book of Mormon), there is good reason to suspect that the actual authors weren't included, and thus good reason to suspect that the results of the study are a probablistic false positive.But, I understand that you have a vested interest in the results since they tend, at least at this point, to support your favored theory. To each their own.Thanks, -Wade Englund- Link to comment
Uncle Dale Posted December 21, 2008 Share Posted December 21, 2008 ...But, I understand that you have a vested interest in the results since they tend, at least at this point, to support your favored theory. To each their own....Well, it did not start out as a theory of mine -- and through the years I've added practically nothing to it, save for some citations of obscure references and a bit of phraseology comparison. I am not so much attached to a "theory" as I am to the proposition that one can be a Latter Day Saint and not have to profess a belief in ancient Nephites.That proposition is still up for debate among the Reorganized LDS, but we seem to have advanced to the point that it does not entail an automatic excommunication.Perhaps future word-print studies will NOT show such a huge probability gap between the Stanford teams results for Spalding/Rigdon, and all the other possible authors tested.So -- if that future 100-author word-print study is carried out, and the gap between the Spalding/Rigdon probabilities and the other tested authors' probabilities is an insignificant one -- then you can point that fact out to me, (and know you've won).We shall see.UD Link to comment
Bender Posted December 21, 2008 Share Posted December 21, 2008 wenglund Posted Today, 01:48 PMI am not speaking about computer error. I am speaking about sample error. If the sample doesn't include the actual author of the text (and there is no way to know if this has occured or not since this is ultimately what is at issue), then the study will invariably return a false positive (meaning, the wrong person will be attributed, probabalistically, as the author).Since this is a long thread you may have missed posts 226 and 227 which I think may answer your concerns about false positives. Link to comment
wenglund Posted December 21, 2008 Share Posted December 21, 2008 Well, it did not start out as a theory of mine -- and through the years I've added practically nothing to it, save for some citations of obscure references and a bit of phraseology comparison. I am not so much attached to a "theory" as I am to the proposition that one can be a Latter Day Saint and not have to profess a belief in ancient Nephites.That proposition is still up for debate among the Reorganized LDS, but we seem to have advanced to the point that it does not entail an automatic excommunication.Perhaps future word-print studies will NOT show such a huge probability gap between the Stanford teams results for Spalding/Rigdon, and all the other possible authors tested.So -- if that future 100-author word-print study is carried out, and the gap between the Spalding/Rigdon probabilities and the other tested authors' probabilities is an insignificant one -- then you can point that fact out to me, (and know you've won).We shall see. UDWere I looking to "win", then I may wait around to point that out. But, I am not. I happen to be of the belief that the origins of the Book of Mormon are only indirectly relevant to the intents and purpose of the restored gospel of Christ--and this as a catalyst for promoting faith in the restored gospel. I believe that the Book of Mormon is ancient in its origins, and this instills trust within me to act upon the precepts contained therein and to trust that modern revelations are also true and beneficial for me. And, having acted on that trust, I coninue to realize the promise of Alma regarding the "swelling of my breast". I have found that the restored gospel of Christ does for me what it claims it will do and what it is designed to do (i.e bring me to Christ), and it has been the best means I have found for becoming the best person possible, and increasing towards a fulness of joy and love.To me, as long as people are utilizing the gospel in the way it is designed and for the purpose it was designed, that's what really matters. If their utilization in this regard is a result of belief in the ancient origins of the Book of Mormon, then great! If their utilization occurs absent that belief, then still great! Thanks, -Wade Englund- Link to comment
Uncle Dale Posted December 21, 2008 Share Posted December 21, 2008 ...If their utilization occurs absent that belief, then still great! ...Bless you, Bro. Englund!UD Link to comment
wenglund Posted December 22, 2008 Share Posted December 22, 2008 My appreciation to Bender for bringing this to my attention.Your supposition is true that if A wrote a book and is not included in the study, whoever is tagged as top author is clearly incorrect. However, that is an immature way of looking at the results in regard to what inverse theory can tell us about complex systems even when useful information is missing. Uncle Dale's response was quite prescient when it comes to interpreting results like these.In short, while a false positive will of course appear in the absence of the actual author, there is a predictable signal of what that should be (Danna, Uncertain, and myself have all mentioned this within this thread). While the result would register a false positive, that false positive would be extremely weak and statistically attributable to chance. E.g., where Longfellow was tagged, it was always weak. Even if one thought Longfellow wrote those particular chapters, it would be an uphill battle to say so with confidence. It is different altogether when a positive registers strongly like the Rigdon and Spalding signals do (and Cowdery and Pratt to a lesser extent).The only way to get consistent strong false positives (i.e., significantly statistically more like one incorrect target author than another, is if that particular author has a strikingly similar voice to the correct one. So one would need to suppose that (in a tight translation paradigm) that Nephi, Alma, Mormon, and others all had strikingly similar voices to Rigdon and/or Spalding relative to other subset of 19th century authors. It is a little easier to fathom that in a loose translation, that the Joseph Smith signature would look far more similar to Rigdon-Spalding than the others, but not by lots.Perhaps you can help my alleged immature way of looking at this by explaining how the study can determine whether a particular author being tested has a strikingly similar voice to "the correct one", or if that particular author is "the correct one"--particularly if "the correct one" is what is in question?And, if there is no way to make that determination, then how can we have confidence that "the correct one" is attribute as the author without presuming the absence of a strikingly similar voices in the study?Thanks, -Wade Englund- Link to comment
NorthboundZax Posted December 22, 2008 Share Posted December 22, 2008 My appreciation to Bender for bringing this to my attention.Perhaps you can help my alleged immature way of looking at this by explaining how the study can determine whether a particular author being tested has a strikingly similar voice to "the correct one", or if that particular author is "the correct one"--particularly if "the correct one" is what is in question?And, if there is no way to make that determination, then how can we have confidence that "the correct one" is attribute as the author without presuming the absence of a strikingly similar voices in the study?This has been answered both here and on MDB, including the section you just quoted. I realize that I have a tendency to write somewhat opaquely, but Danna's post #227 among others addresses exactly this issue. To try to clarify, though, in this kind of measure, with no reason to believe that any particular author has a statistically similar voice to the actual author, no author would be way out in front of the others, just small differences in the probabilities (like in Danna's example). For a voice to register as a strong false positive, that voice must have a statistically similar voice (note that weak first place authors shouldn't really be considered as false positive as they should be discounted as deriving from chance, but for good or bad people are using the terminology as if they are). What this means is that a 'strikingly' similar voice would register as a significant probability even in the presence of the actual author. You could argue that Alma and Spalding just happen to have statistically similar voices that would not be readily statistically distinguishable from each other (i.e., independent writings of Alma should register high on a test of works known to be written by Spalding) - it wouldn't be very parsimonious (especially since the argument would need to be made for other Nephite prophets as well), but it would be a more mature interpretation of the results than your current stance.To me, as long as people are utilizing the gospel in the way it is designed and for the purpose it was designed, that's what really matters. If their utilization in this regard is a result of belief in the ancient origins of the Book of Mormon, then great! If their utilization occurs absent that belief, then still great!I totally agree. Link to comment
Danna Posted December 22, 2008 Share Posted December 22, 2008 And Rigdon would have been in a wonderful position to expose the fraud when Nancy was disignated to be a plural wife and when he, Rigdon, was passed over for the leadership position of the lds church. He would have destroyed the all Brigham church by his declaration: I wrote the book of mormon and here is the manuscript with rough drafts.That assumes that Rigdon viewed the production of the BoM as 'fraud'. On the other hand it was a significant accomplishment for him. And he still believed in the principles he embedded in the 'scripture', so why destroy his own work. It appears that the extent of polygamy had been kept from him, so from his point of view, the over-arching goal of restoration was being achieved - in spite of the weaknesses of the official mouthpiece. After Joseph's death, Rigdon may have been tempted to expose the origin of the book to bring down Brigham, but again, his long-term goal of restoration took priority in the end. He did remain in a pastoral position in a mormon offshoot afterall.I don't think Oliver would have considered the project a 'fraud' either. Again for him, destroying the work in order to avenge himself on Joseph may not have been an obvious option. He may have considered Joseph a fallen prophet, or acting as a 'man'. But likely the enterprise was larger than just Joseph.One of the themes that comes out of the Book of Mormon, and the Book of Abraham as well, is that the end justifies the means. This teleological view is seen in the murder of Laban (and its justification), the theft of the brass plates (along with essentially kidnapping Zoram), and God's instruction to Abraham to lie - among other examples. Like the God of the OT, the God of the Book of Mormon will let innocents suffer in order to instruct or correct a people. Working within that teleological framework, fabricating a new scripture and backstory would be perfectly acceptable in order to bring about the restoration.Check out some of Sidney Rigdon's pre-Mormon writings on Uncle Dale's website. Rigdon was adamant that only the original version of christianity (as defined by him) was acceptable, and restoration of primitive christianity was immense importance to him. Link to comment
Danna Posted December 22, 2008 Share Posted December 22, 2008 ...I happen to be of the belief that the origins of the Book of Mormon are only indirectly relevant to the intents and purpose of the restored gospel of Christ--and this as a catalyst for promoting faith in the restored gospel. ...I think you have neatly summarised Sidney Rigdon's views on the matter! Link to comment
LifeOnaPlate Posted December 22, 2008 Author Share Posted December 22, 2008 That assumes that Rigdon viewed the production of the BoM as 'fraud'. On the other hand it was a significant accomplishment for him. And he still believed in the principles he embedded in the 'scripture', so why destroy his own work. It appears that the extent of polygamy had been kept from him, so from his point of view, the over-arching goal of restoration was being achieved - in spite of the weaknesses of the official mouthpiece. After Joseph's death, Rigdon may have been tempted to expose the origin of the book to bring down Brigham, but again, his long-term goal of restoration took priority in the end. He did remain in a pastoral position in a mormon offshoot afterall.I don't think Oliver would have considered the project a 'fraud' either. Again for him, destroying the work in order to avenge himself on Joseph may not have been an obvious option. He may have considered Joseph a fallen prophet, or acting as a 'man'. But likely the enterprise was larger than just Joseph.One of the themes that comes out of the Book of Mormon, and the Book of Abraham as well, is that the end justifies the means. This teleological view is seen in the murder of Laban (and its justification), the theft of the brass plates (along with essentially kidnapping Zoram), and God's instruction to Abraham to lie - among other examples. Like the God of the OT, the God of the Book of Mormon will let innocents suffer in order to instruct or correct a people. Working within that teleological framework, fabricating a new scripture and backstory would be perfectly acceptable in order to bring about the restoration.Check out some of Sidney Rigdon's pre-Mormon writings on Uncle Dale's website. Rigdon was adamant that only the original version of christianity (as defined by him) was acceptable, and restoration of primitive christianity was immense importance to him.All these arguments (which have their own problems, most of all that they are rather simplistic assertions without much foundation) are thrown to the winds in light of the circumstantial and historical evidence surrounding the coming forth of the Book of Mormon. It really is that simple. But when one begins with the premise (poisoning the well) that Rigdon must have had good reason to lie about the Book of Mormon, especially when the premise is that there ain't no such thing as angels giving books to people, it is easy to dismiss the honest testimony of the participants. Like when Rigdon's daughter described Sidney, who "in the last years of his life called his family together and told them, that as sure as there was a God in heaven, he never had anything to do in getting up the Book of Mormon. And never saw any such thing as a manuscript written by Solomon Spalding." Link to comment
Cold Steel Posted December 23, 2008 Share Posted December 23, 2008 Before we say too much, one way or the other, it would be necessary that we compile Rigdon's later writings, transcribe the contents, and sort them out as to their nature and purpose.His "revelations" seem to reflect his own mindset, namely that everyone between him and the presidency was cast off and rejected of God. I agree that Rigdon would have been far better off to have come clean on any conspiracy to concoct the Book of Mormon. Certainly with the number of people who supposedly were collaborators on the bookâ??the sheer number it would have taken, I would think that at least one would have come clean. But Cowdery, in poor health, started back for Utah and was rebaptized. He said: "That book is true. Sidney Rigdon did not write it; Mr. Spalding did not write it; I wrote it myself as it fell from the lips of the Prophet. It contains the Everlasting Gospel, and came forth to the children of men in fulfillment of the revelations of John, where he says he saw an angel come with the Everlasting Gospel to preach to every nation, kindred, tongue and people."He would have been better off where he was, had it all been a fraud. All it takes is one person in a conspiracy to spill the beans, yet with all those spin-offs and vying for power and hard feelings and conflicting opinions, there was nary a leak of fraud. In fact, the most passionate anyone became in the midst of all this was when it was rumored that one of them had abandoned his testimony. It always drew an immediate, direct and heartfelt rebuttal from each and any of the witnesses. In fact, they've always been one of the most forceful evidences of the book in my opinion. Link to comment
Gervin Posted December 23, 2008 Share Posted December 23, 2008 All it takes is one person in a conspiracy to spill the beans,more often than not, all it takes is one person getting caught, then the beans spill. folks don't just go around spilling beans unless it's to their advantage and if you're part of a conspiracy you better be damn sure you're going to come out on the other side smelling like a rose .. Link to comment
Danna Posted December 23, 2008 Share Posted December 23, 2008 All these arguments (which have their own problems, most of all that they are rather simplistic assertions without much foundation) are thrown to the winds in light of the circumstantial and historical evidence surrounding the coming forth of the Book of Mormon. It really is that simple. But when one begins with the premise (poisoning the well) that Rigdon must have had good reason to lie about the Book of Mormon, especially when the premise is that there ain't no such thing as angels giving books to people, it is easy to dismiss the honest testimony of the participants. Like when Rigdon's daughter described Sidney, who "in the last years of his life called his family together and told them, that as sure as there was a God in heaven, he never had anything to do in getting up the Book of Mormon. And never saw any such thing as a manuscript written by Solomon Spalding."Yes, of course I am speculating - in response to claims along the lines of "if Rigdon really wrote the Book of Mormon he would have ... (insert act of revenge exposing conspiracy to write BoM) ... in response to ...(some offense during his long and varied career)... Link to comment
Benjamin McGuire Posted December 23, 2008 Share Posted December 23, 2008 Just another comment on this study. From the website, they provided a chart showing the groupings from all the texts which they compared (the vocabularies that is) to the Book of Mormon. The closest match was the Isaiah and Malachi material. The next closest authors were Longfellow and Barlow. because of this closeness, we would expect (if the authors were entirely unrelated) that in general Isaiah/Malachi would be preferred most of the time to Longfellow and Barlow (since the Longfellow and Barlow samples are not quite as close to the Book of Mormon samples as the Isaiah/Malachi samples). Had we taken the the Isaiah/Malachi author out of the study (which was never done), most of the Isaiah/Malachi samples would not have become Cowdry or Rigdon or Spaulding samples - they would have become Longfellow and Barlow identifications. This suggests that the low Longfellow and Barlow comparisons are in fact contrived.The studies basis for determining error rates seems to be a bit biased. If we get rid of the glib explanation for why the Isaiah/Malachi results don't really mean anything, the error rate in the study is much higher than the error rate they obtained against known authors. In the original context (cancer studies), this statistical model was built to attempt to get accurate identifications using a small set of samples - when applied across all several thousand genes, cancer types could be determined with one hundred percent accuracy already (using other models). The question was whether or not we could reduce the necessary sampling pool to 100 or even fewer genes to accurately determine the cancer type with the same accuracy. And it was discovered that it was possible - but the model determined the genes which needed to be evaluated.In this case, the genes (the vocabulary words) are chosen through a completely different (and seemingly subjective) process, and so there doesn't seem to me to be the same kind of process involved. The method in the study boils down to the most common shared words (which could actually be replaced by taking the 100 most used words of the time period - to the same essential effect). However this does not seem (intuitively at least) to be able to say a lot about a particular author's choice of vocabulary and its unique features.Ben McGuire Link to comment
wenglund Posted December 23, 2008 Share Posted December 23, 2008 This has been answered both here and on MDB, including the section you just quoted. I realize that I have a tendency to write somewhat opaquely, but Danna's post #227 among others addresses exactly this issue. To try to clarify, though, in this kind of measure, with no reason to believe that any particular author has a statistically similar voice to the actual author, no author would be way out in front of the others, just small differences in the probabilities (like in Danna's example).What reason is there to suppose that a particular author or authors don't have statistically similar voices to the actual author? In fact, doesn't the "small differences in probability" you mention above suggest that there may be statistical similarities in the voices already included in the study?For a voice to register as a strong false positive, that voice must have a statistically similar voice (note that weak first place authors shouldn't really be considered as false positive as they should be discounted as deriving from chance, but for good or bad people are using the terminology as if they are). What this means is that a 'strikingly' similar voice would register as a significant probability even in the presence of the actual author.This explains what may be expected to happen when there is a "statistically similar voice" in the presense of the actual author. However, it doesn't answer my question whether the wordprint techniques can determine if a "significant probability" represents a "statistically similar voice" or the voice of the actual author--and this absent capreciously assuming there is no statistically similar voice, and absent knowing whether the actual author has been included in the study or not? For example, let's assume that author "A" has a statistically similar voice to author "G" (the actual author of the text in question), whereas authors "B" through "D" aren't as statistically similar, and authors "E" and "F" are used as controls and are statistically dissimilar in voice to author "B". And, let's say that all the authors but "G" are included in the study. If I understand the wordprint techniques correctly, the results of this hypothetical study would suggest that there is a significant probability that author "A" is the author of the text in question. The wrong author, then, would be given attribution. Right?What are the chances of this happening with the Stanford study? Who knows? But, if the actual author9s) is/are not included in the study, then there may be as much as 100% chance of a false positive depending upon how statistically similar the voices included in the study are to the actual author(s).In other words, while the wordprint techniques may rule out study results where there is a "weak firstplace author", they can't, in cases where there is a "strong firstplace author", tell us whether the actual author has been included in the study or not, or whether the results of the study are a false positive or not (i.e. whether the study attributes a high probability to the wrong author).Thanks, -Wade Englund- Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.