CaliforniaKid wrote:I would like to try to refocus and to inject a little sanity into this discussion. Uncle Dale has described his methodology, which basically was to make word lists for the Spalding story and for each page of the 1830 Book of Mormon and to determine what the % vocabulary overlap is. If I understand him correctly, anywhere there was an overlap of more than 95 %, he colored it red.
You've got part of that observation right, CK -- but not all of it.
The red-colored sections on the charts were determined by a rough quantification of word-strings, not by
an undifferentiated shared vocabulary count. That is why I posted the 3-part chart (Alma XX-Helaman I)
earlier in this thread. I'll repost it now, in order to clarify things.
This chart of the "Book of Solomon" has been reduced in size to fit easily into this web-page,
but the full-sized original is also available for viewing here:
http://solomonspalding.com/SRP/MEDIA/phrchrt2.gif The vocabulary count you are speaking of, CK, is indicated by the upper graph line, which roughly varies
from a low of about 90% (on p. 395) to a high of about 98.2% (374, 385) in the pages of this section of the
1830 Book of Mormon. In other sections of the book, that upper graph line dips down to as low as 80-85% if I recall correctly.
I do not have all the data available to me right now, as another person is working with this stuff, helping me
with analysis, etc.
The red columns are what you are talking about, and on the above chart they depict the level of fine detail
missing from my highly generalized 1980 JWHA paper chart, (where red colored sections represent my
conclusions regarding "Spaldingish" Book of Mormon text).
The red column data, as I said, were determined by a rough quantification of word-string count per 1830 Book of Mormon
page. I charted out the occurrence of 515 shared word-strings, and wherever I found 16 or more on a Book of Mormon page,
I colored the column for that page red on the chart. Pages containing 11 to 15 such shared strings I colored
orange, representing "undecided" for my evaluation purposes. Pages containing 0-10 such strings are green,
representing pages which I have subjectively determined to be "un-Spaldingish."
Thus, if a page like #340 (the first page on the 3-part chart) has in its first half, no shared word-strings, and in
its second half 10 shared word-strings, that section of the text is marked green ("un-Spaldingish") even though
a portion of its contents overlaps greatly with the language of the Oberlin manuscript. So, just because a column
has a certain color on the graph, that does not mean the text on the page thus represented is homogeneous. This
factor must be kept in mind when we observe the juxtapositioning of the upper line graph over the data shown
in the bottom column graph. The data thus depicted in those two parts do not align in perfect sync -- partly
because certain pages are not homogeneous in their "voices."
At any rate, I think that the upper line graph and the bottom column graph correlate fairly well. We are are
observing only a small section of the overall Book of Mormon text -- but the two graphic indicators I've been talking about
here are at their "mountain peaks" level in this section of the 1830 book. In most other, very lengthy sections
of the Book of Mormon text, the bottom column chart count is in the "green range" and the upper line graph is fairly flat,
at the bottom of its range.
In between the two graphic indicators we have just been looking at, is a third indicator -- the heavy blue line.
This represents the Solomon Spalding non-contextual word-markers count (or word-print) for ranges of text
segments in the Book of Mormon (generally more than a single page, because word-printing requires larger text segments).
The ups and downs of this blue line have been purposely flattened in my chart (at the request of the person
who gave me that data) so as not to divulge pre-published information scheduled for an upcoming professional
paper. However, I will say here and now that the Spalding word-print, as charted throughout the Book of Mormon by an
independent researcher, also has a "close fit" with the patterns you see in my own two graphic indicators.
That is why I added the blue word-print line graph to my own depictions, CK. The length you can see on my
3-part chart represents the Himalayas of that graph line, when extended across the entire Book of Mormon text.
Now, let me digress on the subject of Metcalfe's lexical shift for a moment... It's only because a pattern
emerges that Metcalfe's data are meaningful.
Now, Uncle Dale's choice of 95% as a cutoff point is fairly arbitrary. At this point, it isn't meaningful that some pages have a 95% vocabulary overlap. What is meaningful, in my mind, is that it works. The pages with 95+% overlap cluster in the book of Alma, and especially in the second half of the book of Alma. A pattern emerges. Why?
Since you have made an incorrect assumption about the chart data, CK, I will correct you here and say that
the 16+ word-string count is the "cutoff point" you really mean to be speaking of here. I arbitrarily selected
515 shared Spalding/Book of Mormon word-strings, which I felt might be "significant" in terms of the number of words thus
connected in each string, and in terms of their thematic content. I challenged Dan Vogel to select his own two
sets of roughly 500 shared word-strings, and to chart out his data in a similar manner. My prediction is that any
two charts he might produce by this method will roughly duplicate the patterns I have shown in my chart.
The "why" is the important part. Uncle Dale's black-and-white chart mapping out "spikes" in vocabulary overlap coinciding wih Book of Mormon battles and religious material offers one possible explanation for the "why." Do these pages deal with some subject matter that frequently occurs in the Spalding manuscript? Does a coincidence of topic cause a superficial coincidence in vocabulary that clusters in these pages? Or does the coincidence in vocabulary mean something else, namely that these pages were authored by the same person who authored the Spalding manuscript?
Again I must correct you, CK --- since we are talking about the arbitrarily selected 515 word-strings in the
red bar graph (and about 1500 other arbitrarily selected word-strings in b&w Oberlin MS graph), the "spikes"
which thus show up are a direct outcome of that arbitrary selection process.
Please read carefully here -----> This is why the line graphs (which are not based upon arbitrarily selected data)
serve as important controls on the bottom column chart. The column chart is subjective stuff, the line graphs are
objective stuff.
Uncle Dale can strengthen the case for the latter by pursuing the following steps:
1) Choose several other, similar but independent 19th-century texts to serve as controls. Perhaps Dan can help you select them. The texts you choose should probably be written in mock-archaic or biblical prose, maybe along the lines of the Third Epistle of Peter, since the Book of Mormon is written in said prose.
2) Apply the same method to these texts that you applied to the Book of Mormon: a page-by-page comparison of vocabulary overlap.
3) If a cluster of 95+% vocab overlap occurs in any of your control books, then your 3-part chart probably tells us more about a coincidence of subject matter than about authorship. If no such overlap clusters appear, then your chart may in fact have implications for authorship attribution.
If we can get Dan to agree that this is an acceptable "way forward," I think I can write a fairly simple Visual Basic program to computerize and facilitate the vocab comparison process, so long as we have accurate typescripts of each work we are comparing. Or perhaps, since Brent is a software engineer and has shown an interest in laying the Spalding theory to rest, he would be interested in writing such a program.
-CK
I think Brent's only interest is in telling folks not to waste 10 seconds on Spalding/Rigdon theories. I predict he
would not touch such material with a ten-foot pole, and would only browse through the results, looking at the
"more part" of the findings, in order to locate some small gnat of a perceived problem, which he might blow out
of all reasonable proportion -- so as to obscure the general message of those particular findings.
Now Dan may, in this case, again say I am attacking the messenger, and not listening to his message. But I
will be very, very surprised if Brent will agree to spend any time at all upon such a project.
As for texts to examine, as controls, look here for some possible selections:
http://olivercowdery.com/texts/bookindx.htm
Cheers,
UD