ERDNASE

[mam]

On the other hand, finding documentation of any of the candidates being member of any of the clubs wouldn't really prove anything. There were clubs. They had members.

[lybrary]

I both agree and disagree here. Yes, knowing that one of the candidates was a member at certain clubs would not prove anything, unless there is a card players club. But knowing who else was member could allow us to better understand the connections and friendships a candidate had.

For example, we know that Edward Gallaway was a member of the OddFellows. I know that he joined them in summer of 1902. I am now trying to get the full membership roster of his lodge, Excelsior no. 22, to see who else was member there. This could explain parts of what we know about EATCT, where it was advertised and sold. Imagine for example - and that is purely hypothetical - that Vernelo was also member of the Excelsior lodge. That could then explain why EATCT appears advertised in the Sphinx later in 1902. As a self-published author Erdnase has to try to market and sell his book. He needs to find ways to sell the print run he paid for.

It is these connections which could be very helpful to explain certain things known to us.

[Brad Jeffers]

D. B. Cooper comes to mind

Speaking of which ...

Would you believe me if I told you that I have uncovered an edition of The Expert At The Card Table which is signed and inscribed by the author, and which at one time belonged to the hijacker D.B. Cooper?

Of course you wouldn't.

But it's true.

[lybrary]

D. B. Cooper comes to mind

Speaking of which ...

Would you believe me if I told you that I have uncovered an edition of The Expert At The Card Table which is signed and inscribed by the author, and which at one time belonged to the hijacker D.B. Cooper?

Of course you wouldn't.

But it's true.

Brad, so you are saying you have a handwriting sample of Erdnase?

[Bill Marquardt]

D. B. Cooper comes to mind

Speaking of which ...

Would you believe me if I told you that I have uncovered an edition of The Expert At The Card Table which is signed and inscribed by the author, and which at one time belonged to the hijacker D.B. Cooper?

Of course you wouldn't.

But it's true.

I will take your word for it.

My first encounter with EATCT was a mail order copy of an early paperback edition that I purchased for $3.00 circa 1961. Many years later it ended up in a box in an outside storage shed and was chewed to pieces by mice. I threw away the remains without much concern as I knew the book was readily available. I recently saw a similar edition offered at a generous price on eBay, nothing that would have made me rich had I still owned a copy to sell, but I am disgusted that I let my copy go into outside storage. Call me a numbskull.

My book was not signed, however. Except by me.

[mam]

I found this book from 1888 that has membership lists of a number of clubs in Chicago. A quick scan says Gallaway, Drake, McKinney or Ruxton are none of them members of any of those clubs. But feel free to look closer at the lists.

[Joe Pecore]

Everybody take a well deserved break from researching and relax to the jazz styling of the Janek Gwizdala Band playing the song "Erdnase"



Check out Janek (amatuer magician and bassist) Gwizdala's other songs on his album "Theatre By The Sea" https://janekgwizdala.bandcamp.com/albu ... by-the-sea , which includes other magical one like "Fooling Houdini", "The Goshman", and "Chicago Opener".

[Bob Coyne]

This is my announced quantitative analysis. It is my first attempt putting some numbers behind some of the evidence that gets mentioned in lists. It is my first stab and I welcome critique, comments, and suggestions to make it better.

Chris, there's something I don't understand about your reasoning on this. Your starting assumptions (which seemed valid to me) were that erdnase was engaged directly (or possibly indirectly) in a business relationship with mckinney to print the book. However, I don't understand why Gallaway would count as a business relationship of the sort that matters. Is it just because he's listed as a creditor on the bankruptcy files? It seems likely that's just because he was an employee and was owed money for back pay or some other reason related to his role as an employee as opposed to that being evidence that he was a customer (as erdnase would have been). At a minimum the relatively likelihood of those two scenarios (being a creditor as a customer vs as an employee or in some other way) must be taken into account in your calculations. It's the same reason you can reduce/eliminate the likelihood of others on the creditors list (eg cook county, dexter folding company, etc etc). The key question isn't whether someone was a in a business relationship but whether they were in a business relationship because they were a customer as an author to get the book printed.

[lybrary]

This is my announced quantitative analysis. It is my first attempt putting some numbers behind some of the evidence that gets mentioned in lists. It is my first stab and I welcome critique, comments, and suggestions to make it better.

Chris, there's something I don't understand about your reasoning on this. Your starting assumptions (which seemed valid to me) were that erdnase was engaged directly (or possibly indirectly) in a business relationship with mckinney to print the book. However, I don't understand why Gallaway would count as a business relationship of the sort that matters. Is it just because he's listed as a creditor on the bankruptcy files? It seems likely that's just because he was an employee and was owed money for back pay or some other reason related to his role as an employee as opposed to that being evidence that he was a customer (as erdnase would have been). At a minimum the relatively likelihood of those two scenarios (being a creditor as a customer vs as an employee or in some other way) must be taken into account in your calculations. It's the same reason you can reduce/eliminate the likelihood of others on the creditors list (eg cook county, dexter folding company, etc etc). The key question isn't whether someone was a creditor but whether they were a creditor because they were a customer as an author to get the book printed.

Bob, I am simply taking a group of people of which Erdnase must be member of. This is of course an upper bound, and there are probably several ones which could be ruled out. But clearly, being an employee as Gallaway was, gives you all the opportunity to have your book printed at your workplace. Keep in mind that Gallaway wasn't some lowly printing laborer. He was 20 years in the business, advanced quickly through the rungs of his profession and must have held a higher position at McKinney. We see this from his wage as well as from other sources of information who show that he was a clever, ambitious, and successful person.

[Brad Jeffers]

D. B. Cooper comes to mind

Speaking of which ...
Would you believe me if I told you that I have uncovered an edition of The Expert At The Card Table which is signed and inscribed by the author, and which at one time belonged to the hijacker D.B. Cooper?
Of course you wouldn't.
But it's true.

I will take your word for it.

Don't just take my word for it Bill, it's true.
Let me explain ...

The particular edition to which I refer is this edition ...



It is inscribed ...
"To Bill Gossett ~ Remember Gamblers Never Gamble ~ Michael MacDougall"

William Gossett is believed by many to be the legendary D.B. Cooper.

Therefore when I say I have an edition of The Expert At The Card Table that is signed and inscribed by the author and which at one time belonged to the hijacker D.B. Cooper, you can believe it!

[Bill Marquardt]

@Brad

Very cool.

[Pete McCabe]

https://xkcd.com/1400/

[lybrary]

As a comparison take for example WE Sanders. What do we know about him? Well we know he was an adult male who played cards. My guess for the number of adult male who played cards in the US is about 5 million. What else do we know about him that allows us to narrow it down statistically? We don't even know he was in Chicago at the right time let alone had any business with McKinney. He doesn't sound anything like Erdnase, which is a big point against him.

There are many strong connections with Sanders. First off, Sanders not only played cards but wrote down a magic trick! And he brought several boxes of playing cards with him on a packing list for a trip. Not to mention the Erdnase/earth-nose/mining connection, the history of playing around with anagrams, interest in dialectical speech patterns, and several other strong links.

And I *totally* disagree that Sanders sounds nothing like Erdnase. Unlike Galloway his writing style and word choice is very similar to Erdnase. To me this is one of the strongest of the many pieces of evidence for Sanders. Here's a sampling (i posted many of these sometime ago in this same thread):

Erdnase: he coolly proposes to "MAKE GOOD" by transforming the wrong card
Sanders: Has "MADE GOOD" at the bar, where he shines

Erdnase: It is almost AN AXIOM that a novice will win his first stake.
Sanders: this latter is AN AXIOM in mining during this period of development, and should be invariably followed where possible.

Erdnase: it may enable the skilled in deception to take A POST-GRADUATE COURSE in the highest and most artistic branches of his vocation
Sanders: during the following year he took A POST-GRADUATE COURSE in Civil Engineering

Erdnase: the average card player
Sanders: the average mining engineer

Erdnase: The first described is AN EXCELLENT ONE for retaining either the top or bottom stock...
Erdnase: The position is AN EXCELLENT ONE for ordinary dealing, and should never be changed.
Sanders: this joint is without doubt AN EXCELLENT ONE when, and only when, ....

Erdnase: DESCRIBING with detail and illustration EVERY KNOWN expedient, manoeuvre and strategm of the expert card handler
Sanders: the mines operated under these methods PRESENT EVERY KNOWN characteristic of lode formation.

Erdnase: LITTLE OR NO skill is required, BUT a practiced hand can locate and bring the cards to the top
Sanders: large excavations may be supported with LITTLE OR NO timbering, BUT usually...

Erdnase: various METHODS OF LOCATING AND PRODUCING selected cards
Sanders: the METHODS OF LOCATING AND ALIGNING the sets are those used for...

Erdase: An expert can run the whole deck WITH THE UTMOST rapidity
Sanders: huge timbers that have been framed WITH THE UTMOST precision

Erdnase: though this method IS now BY FAR THE MORE prevalent among men who play for money
Sanders: this station, while requiring more excavating to construct, IS BY FAR THE MOST economical in the end

Erdnase: we shall describe several of the BEST METHODS known for secretly exchanging
Erdnase: the BEST AND SIMPLEST METHODS of accomplishing the sleights
Sanders: probably the SIMPLEST METHOD OF aligning the side plates of inclined-shaft sets
Sanders: being the SIMPLEST AND CHEAPEST METHOD OF framing

Erdnase: if requested to determine from what single artifice THE GREATEST ADVANTAGE is derived we would unhesitatingly decide...
Sanders: the plan above described may be of THE GREATEST ADVANTAGE in blocking-out the ores for purposes of description and localization; and it may be employed with great benefit in connection with...

** Both OFFER a TREATISE and stress the IMPORTANCE of DETAILS **

Erdnase: A TREATISE on the Science and Art of Manipulating Cards
Erdnase: the sum of our present knowledge is PROFFERED THIS IN VOLUME
Erdnase: IMPORTANCE of DETAILS [full section heading]

Sanders: it has appeared worth while to make the present COLLECTION WHICH IS OFFERED not as a complete TREATISE on the subject, but rather as a series of essays which go fully into many IMPORTANT DETAILS

** Both justify the use of CERTAIN TERMS AND SYMBOLS for THE SAKE OF BREVITY **

Erdnase: Many of the methods of card manipulation explained in this work originated with us, and we have, in describing the various processes and conditions, used CERTAIN TERMS for the SAKE OF BREVITY, to DESIGNATE the particular matters referred to.
Sanders: for the SAKE OF BREVITY in description, CERTAIN SYMBOLS letters or figures, are employed to DESIGNATE the various mine workings, as follows:
Sanders: they are thus marked, CERTAIN SYMBOLS may be discarded for the SAKE OF BREVITY, and only such as are essential to the DESCRIPTION of the working be employed.

** Both take time to describe the relevance of the illustrations **

Erdnase: Therefore the writer has expended much time and care in illustrating many manoeuvres that at first may seem unimportant, but all of which are essential to the curriculum of artistic card handling.
Sanders: in the figures drawn to illustrate the article, sizes of timber most frequently used have been arbitrarily taken for convenience. The figures giving dimensions are working drawings showing the methods of framing, as explained, and can easily be applied to frames and timbers of any desired dimensions.

** Both give disclaimers, describing the author's intentions and the limitations of what is covered **

Erdnase: IT IS NOT OUR PURPOSE TO DESCRIBE the various kinds of apparatus, or prepared or mechanical cards, that play so great a part in the professional conjurer's startling exhibitions.
Sanders: IT IS NOT THE PROVINCE OF THIS ARTICLE TO TOUCH UPON methods of mining in use above ground, whether by hydraulic mining, or other processes, but rather to deal with the support of underground excavations by the use of timbers, and the details of mining therewith connected. NOR IS IT INTENDED TO ...

There are more than 130 unique four-word strings matching between ‘Estimating for Printers’ and ‘Expert at the Card Table’. Many appear multiple times in both books

A
• a great deal of
• a manner that the
• a matter of fact
• and as a rule
• and bottom of the
• and it is this
• and so on these
• and so on until
• and there is no
• an inch of space
• another form of the
• any of the other
• a part of the
• are essential to the
• are found in the
• as a matter of
• attention to the fact
• at the bottom of
• at the same time
• at the top and
• at top bottom and
B
• be made in the
• be more or less
• be pressed against the
• be put in the
• both sides of the
• by the use of
C
• cards can be readily
• can be done in
• can be obtained from
• color of the ink
• could be so imposed
• counting the number of
D
• during the process of
E
• end for end and
F
• first and so on
• for all practical purposes
• for the purpose of
• from right to left
H
• his knowledge of the
I
• in about the same
• in addition to the
• inch at the side
• in such a manner
• in the direction of
• in the same manner
• in this particular case
• is not the same
• is one of the
• is placed on the
• is placed on top
• is quite possible to
• is ready for the
• is taken off the
• is the fact that
• is the process of
• it can be done
• it is an excellent
• it is desired to
• it is impossible to
• it is necessary to
• it is one of
• it is quite possible
• it is very important
• it is well to
• it will be seen
M
• manner in which the
• matter of fact the
• merely to show the
• more or less than
N
• number of points to
O
• of course it is
• of course it must
• of the face of
• of the first and
• of the lower one
• of the most important
• of the nature of
• one of the most
• one of the very
• one side and the
• on the other side
• on top of the
P
• particular attention to the
• placed on top of
Q
• quite possible to get
R
• relative positions of the
S
• same result can be
• should be made in
• so on until all
• so that it will
• such a manner that
T
• than the number of
• that are to be
• that it is the
• that it can be
• that it will be
• the back of the
• the bottom of the
• the cards can be
• the center of the
• the color of the
• the face of the
• the face with the
• the first and so
• the most favorable conditions
• the nature of the
• the number of points
• the performance of the
• the pressure of the
• the process is very
• the purpose of this
• there are no more
• the relative positions of
• the same manner as
• the same result can
• the stock must be
• the time required for
• the top and bottom
• the top of the
• the value of the
• the width of the
• to a great extent
• to ascertain the number
• to determine the number
• to go through the
• top and bottom of
• top of the first
• to the back of
• to the fact that
• to the number of
U
• used in connection with
W
• where there is a
• which are essential to
• which will be the
• will be seen that
• with the exception of
• would be required to
• would have to be
• would indicate that the

Who still doubts that these are the same authors?

[Jonathan Townsend]

Chris, how does that match of strings compare to findings from/between other texts of the time?

[Bob Coyne]

There are more than 130 unique four-word strings matching between ‘Estimating for Printers’ and ‘Expert at the Card Table’. Many appear multiple times in both books

[...]

Who still doubts that these are the same authors?

I do A large number of those are too short or generic (eg "a part of the" or "any of the other" "color of the ink" etc) to indicate anything much. Also, sometimes inexact matches that retain some higher level syntactic idiom can be more convincing even though they won't show up on an automatically generated exact match list of this type. Do you have a list of ones that actually sound characteristic of Erdnase to you? I started looking through Estimating for Printers and did run across some language that reminded me of Erdnase, but some other language seemed quite different. So I think it's worth investigating. It takes a while to get familiar with an author's voice. Automated lists is no substitute for that, though i think it can be a useful tool.

[lybrary]

Chris, how does that match of strings compare to findings from/between other texts of the time?

Jonathan, each such comparison, which includes other aspects, too, not just longest substring matches, costs $900. If you are able to raise the funds I am happy to have as many books analyzed as you want.

[Bob Coyne]

Bob, I am simply taking a group of people of which Erdnase must be member of. This is of course an upper bound, and there are probably several ones which could be ruled out. But clearly, being an employee as Gallaway was, gives you all the opportunity to have your book printed at your workplace. Keep in mind that Gallaway wasn't some lowly printing laborer. He was 20 years in the business, advanced quickly through the rungs of his profession and must have held a higher position at McKinney. We see this from his wage as well as from other sources of information who show that he was a clever, ambitious, and successful person.

It seems to me that anybody who lived in or visited chicago would have the opportunity to get the book printed by mckinney. I don't see why is it much more likely for an employee to print a book there than any other person in the vicinity. It's hard to know what gallaway's name on the creditors list means given that he was an employee and could be owed money for any number of reasons. His name on the list increases his likelihood vs a random person but not nearly enough to count him for sure in the estimated 300 people doing business directly with mckinney that year.

[Bill Mullins]

If MS Word's word count is to be believed, EATCT has something like 50,000 words. So it's got something like 49,997 four-word strings. (no idea how many of them are unique). The Gallaway book looks to be about half as big, so it's probably got something like 25,000 words. Take any two books of comparable sizes, and they are bound to have a number of four-word strings in common.

Without some controls, and comparisons to other pairs of books of similar sizes, there's no way to draw any conclusions. Ideally, you'd need accurate word counts on both books and several other books of similar sizes for comparisons. You'd need to build an array of every four-word phrase in each book. Sort the phrases alphabetically so you can easily compare one book to another. If you find that comparing a random 50k word book with a random 25k word book (or whatever the numbers are) yields between 100 and 150 matches, then the comparison between Erdnase and Gallaway shows that there isn't anything unusual about how similar they are (by this metric). If, on the other hand, you usually get something like 50 - 100 matches, then maybe you are on to something. But to say that there are 130 four-word phrases in common, without some comparison to other books, doesn't tell us anything.

[Bill Mullins]

Chris, how does that match of strings compare to findings from/between other texts of the time?

Jonathan, each such comparison, which includes other aspects, too, not just longest substring matches, costs $900. If you are able to raise the funds I am happy to have as many books analyzed as you want.

I had assumed, that with your technical background, you had written a program to do this analysis.

Of the first four phrases on Chris's list, two of them show up with some regularity in other books as well:

[Richard Hatch]

It's hard to know what gallaway's name on the creditors list means given that he was an employee and could be owed money for any number of reasons. His name on the list increases his likelihood vs a random person but not nearly enough to count him for sure in the estimated 300 people doing business directly with mckinney that year.

The money McKinney owed Gallaway is specifically identified as "wages". The checks McKinney issued to his employees on Friday, December 19, 1902 bounced.

[lybrary]

If MS Word's word count is to be believed, EATCT has something like 50,000 words. So it's got something like 49,997 four-word strings. (no idea how many of them are unique). The Gallaway book looks to be about half as big, so it's probably got something like 25,000 words. Take any two books of comparable sizes, and they are bound to have a number of four-word strings in common.

Without some controls, and comparisons to other pairs of books of similar sizes, there's no way to draw any conclusions. Ideally, you'd need accurate word counts on both books and several other books of similar sizes for comparisons. You'd need to build an array of every four-word phrase in each book. Sort the phrases alphabetically so you can easily compare one book to another. If you find that comparing a random 50k word book with a random 25k word book (or whatever the numbers are) yields between 100 and 150 matches, then the comparison between Erdnase and Gallaway shows that there isn't anything unusual about how similar they are (by this metric). If, on the other hand, you usually get something like 50 - 100 matches, then maybe you are on to something. But to say that there are 130 four-word phrases in common, without some comparison to other books, doesn't tell us anything.

Bill, not according to the expert who did the analysis. The substring match was only one thing he looked at, but according to him, it is one very significant metric. I would assume he knows what he is talking about. Another thing he looked at was the common lexicon. Here is what he writes:

"the common lexicon and the high number of four word phrases suggests a strong possibility of identity of authorship"

[lybrary]

Of the first four phrases on Chris's list, two of them show up with some regularity in other books as well

Which means two of the four you picked at random are quite significant. Based on your sample we can assume that perhaps 50% or about 60-70 are uncommon. I don't want to bore you with another statistical analysis, but 130 matches many of which are used repeatedly is highly significant. The WE Sanders camp should raise the money for an analysis for Sanders. Then you would have an argument to stand on. Right now it is you against an expert who does this all the time. I believe the expert if you don't mind

[Jonathan Townsend]

Significance and confidence. How many four word fragment matches is "normal" and what are the odds that a finding is due to chance alone.

[lybrary]

Significance and confidence. How many four word fragment matches is "normal" and what are the odds that a finding is due to chance alone.

Jonathan, as I wrote, this is only one metric the expert used to come to his conclusions. It is not as simple as looking at a number and saying it is high or low.

I offered the data as comparison to what Bob Coyne put up for Sanders which are mostly three word phrases several of which don't even match exactly.

[DChung]

Jonathan, each such comparison, which includes other aspects, too, not just longest substring matches, costs $900. If you are able to raise the funds I am happy to have as many books analyzed as you want.

I hope those other metrics are pretty good, because otherwise you are seriously getting ripped off.

Here's one I found online written in C++ http://www.siafoo.net/snippet/380
Ok, it's pretty terrible code, but if know just a little C++, you can easily tweak it to get what you need, and Simply making sure to set limits on the size of the substrings to check should be enough to make the thing run quickly enough.

That being said, if anybody does want to pay $900, I can easily find some poor student (or just some bored programmer) to write the program from scratch for you, and that's for a program that will match any two inputs you throw a it.

Also, I feel the need to point out that I am unconvinced by Chris's analysis. Bill's point of comparing other books is absolutely important. You can't draw conclusions by just running "experiments" on your own candidate. This seems to be a crucial mistake that Chris isn't the only one guilty of on this thread. I've seen it throughout the long conversation and I imagine many of those same people will likely dismiss what I have to say about the matter as it doesn't bolster their case.

If I found Erdnase's DNA and showed that it matched AT LEAST 99% with someone alive today, that could sound convincing that I've found a descendent or relative. BUT if you know your biology, all humans share 99.5% of their DNA, so it actually means absolutely nothing. If you're going to convince somebody that a particular characteristic is DISTINGUISHING, then you have to show that it's NOT a common trait. Otherwise, the only person you're convincing is yourself.

Best,
Derrick

P.S. Chris, it's possible that you do have more convincing evidence about this given by your expert. All I have to judge on is what you've shared here, and in my eyes, that evidence is wholly unsatisfactory. Also you seem to be bandying the word significant around quite a bit. From what's been written on this board, I can see nothing to suggest that any of these matches are significant in any quantitative way.

[lybrary]

Also, I feel the need to point out that I am unconvinced by Chris's analysis. Bill's point of comparing other books is absolutely important. You can't draw conclusions by just running "experiments" on your own candidate. This seems to be a crucial mistake that Chris isn't the only one guilty of on this thread. I've seen it throughout the long conversation and I imagine many of those same people will likely dismiss what I have to say about the matter as it doesn't bolster their case.

Except you are forgetting that the expert has made such comparisons many times before and thus is very well aware of what is to be expected and what not. That is exactly why I have hired an expert. I myself, just as you and probably everybody else here, does not know what is significant and what is not. The forensic linguist says it is significant.

But I will wait for your analysis of a dozen other books since you believe it is so easy. I have put up my own money for my candidate. Now I want to see those who are quick with criticism, which is based on zero experience in forensic linguistics, to put their money where their mouth is. I am more than happy to compare the linguistic fingerprint of Gallaway against any other candidate.

[Bill Mullins]

I'm not "against" the expert. I'm simply saying that the statement that "130 matches in works of this size is significant" can't be evaluated except in comparison to how many matches one would expect from comparing other similar works. Otherwise we are accepting an "argument from authority" (without even knowing who the authority is), which is the same as accepting Pratt's statement that Erdnase was MFA.

The only way I can interpret what your expert is saying is that 130 is higher than one would expect if the two works were not written by the same person. But how much would one expect? If one would only expect 125 to be in common, that's not all that significant. If one would expect 20 or 30, this would be very significant. So what is the threshold? And how is it determined? That's the thing about analysis --- it should be repeatable. Given your expert's assumptions and methodology, anyone should be able to repeat what he has done, and come to the same conclusion. We don't know his assumptions, and we only know part of his methodology.

I agree that a common lexicon would be significant. That was the basis of my post of a few days ago, where I looked at words/phrases that appear regularly in either Erdnase or Gallaway, but not in the other. I showed that there are several lexemes that are common within either Erdnase or Gallaway, but not so common within the other author's writing. Does your expert have an opinion on that issue?

Based on your sample we can assume that perhaps 50% or about 60-70 are uncommon.

No, we can't. Because we don't know if the other two appear in other works as well. I simply observed that two of the phrases were significantly more common than the other two, and surmised that they must appear in more places. I didn't say (and cannot say) that the two scarcer phrases appear only in Erdnase and Gallaway, or that they are "uncommon". Further, my sample isn't random (it was simply the first four in alphabetical order, and the alphabet isn't normally distributed and represented in English), so you can't say that two out of four applies to the whole list.

[lybrary]

Bill, you want to reduce the knowledge of an expert to a number. That will not work. I suggest you read one of the textbooks on forensic linguistics: http://www.lybrary.com/forensic-linguis ... 04427.html

The analysis for Gallaway was prepared by Dr. Olsson.

[Bob Coyne]

Significance and confidence. How many four word fragment matches is "normal" and what are the odds that a finding is due to chance alone.

Jonathan, as I wrote, this is only one metric the expert used to come to his conclusions. It is not as simple as looking at a number and saying it is high or low.

I offered the data as comparison to what Bob Coyne put up for Sanders which are mostly three word phrases several of which don't even match exactly.

Matching exactly isn't the issue, It's how characteristic the phrases are. Also, the examples I gave weren't mostly three word phrases. Instead they were sentences and other longer phrases where Erdnase and Sanders actually sound very much alike (to me at least -- everyone can judge for themselves).

As couple examples of larger patterns that don't match word-for-word but ring out much more than generic four word phrases like "a part of the" in your list:

Erdnase: we have, in describing the various processes and conditions, used CERTAIN TERMS for the SAKE OF BREVITY, to DESIGNATE the PARTICULAR matters referred to.
Sanders: for the SAKE OF BREVITY in description, CERTAIN SYMBOLS letters or figures, are employed to DESIGNATE the VARIOUS mine workings, as follows:

Erdnase: It is an excellent manner of holding the deck for the true shuffle, and SHOULD BE STRICTLY ADHERED TO ON ALL OCCASIONS.
Sanders: this latter is an axiom in mining during this period of development, and SHOULD BE INVARIABLY FOLLOWED WHERE POSSIBLE.

[lybrary]

Erdnase: It is an excellent manner of holding the deck for the true shuffle, and SHOULD BE STRICTLY ADHERED TO ON ALL OCCASIONS.
Sanders: this latter is an axiom in mining during this period of development, and SHOULD BE INVARIABLY FOLLOWED WHERE POSSIBLE.

Do you mean this seriously? Or are you kidding? It is neither a match nor does it mean the same.

[DChung]

Chris,

My statement was based on the evidence provided here on this forum. You may have plenty of solid evidence from John Olsson but it has NOT been presented here.

Also, I'm really just responding because you are using "bad math" to try and support your case. Avoid that, and I won't feel the need to chime in at all. I'd recommend Darrell Huff's "How to Lie with Statistics" as something to read before jumping to conclusions with any statistical data you have.

I DID say that finding common substrings between two texts is relatively easy to do. That is all. I have no idea what other analysis was done, but I imagine that has to be where the meat is. I just found it strange that what you presented was the substring analysis, because certainly without comparison to other books that evidence is weak and incomplete. I trust that the other evidence is stronger, but you haven't shared it here.

Now it could be that any other book compared to Erdnase has FAR less similarities, but that's something that has to be checked. Now it's again likely that your expert checked this or knows the numbers, but it's not something you shared here. I'm curious whether the people who've bought your book feel differently from me.

And I've got no candidate, so I'll keep my 900 bucks. Just saying that as an interested impartial observer, from what I've read on this forum, I'm not convinced.

Derrick

[Bob Coyne]

Erdnase: It is an excellent manner of holding the deck for the true shuffle, and SHOULD BE STRICTLY ADHERED TO ON ALL OCCASIONS.
Sanders: this latter is an axiom in mining during this period of development, and SHOULD BE INVARIABLY FOLLOWED WHERE POSSIBLE.

Do you mean this seriously? Or are you kidding? It is neither a match nor does it mean the same.

Of course! You seem to be stuck on the concept of direct matches versus similarity, particularly in voice/style. Are you a native English speaker?

"Should be" = "should be"
"strictly adhered to" = "followed" (hint "strictly" is an adverb, modifying the head words "adhere to" which is synonymous with to "follow")
"on all occasions" and "where possible" are both modifiers removing limits on when the adhering/following should take place. The fact that the action is qualified is part of the similarity.

Of course I could also mention how "is an axiom" in sanders here also echoes erdnase elsewhere with "It is almost AN AXIOM that a novice"

And since you seem to have had such trouble with this one I'll map out the other one for you:

Erdnase: we have, in DESCRIBING the various processes and conditions, used CERTAIN TERMS for the SAKE OF BREVITY, to DESIGNATE the PARTICULAR matters referred to.
Sanders: for the SAKE OF BREVITY in DESCRIPTION, CERTAIN SYMBOLS letters or figures, are employed to DESIGNATE the VARIOUS mine workings, as follows:

describing = description (nominal form vs verbal form)
Designate = designate
"Certain terms" = "Certain Symbols" (don't get hung up on terms vs symbols...it's all in the context of the illustrations and their utility)
"for the sake of brevity" = "for the sake of brevity"

In both cases the authors JUSTIFY (that word isn't there...but it's ok, that's just part of how the authors attitude/voice coming through implicitly) the use of specific terms and symbols in the illustrations of the books they wrote.

..and there are many more.

[lybrary]

Chris,

My statement was based on the evidence provided here on this forum. You may have plenty of solid evidence from John Olsson but it has NOT been presented here.

Also, I'm really just responding because you are using "bad math" to try and support your case. Avoid that, and I won't feel the need to chime in at all. I'd recommend Darrell Huff's "How to Lie with Statistics" as something to read before jumping to conclusions with any statistical data you have.

I DID say that finding common substrings between two texts is relatively easy to do. That is all. I have no idea what other analysis was done, but I imagine that has to be where the meat is. I just found it strange that what you presented was the substring analysis, because certainly without comparison to other books that evidence is weak and incomplete. I trust that the other evidence is stronger, but you haven't shared it here.

Now it could be that any other book compared to Erdnase has FAR less similarities, but that's something that has to be checked. Now it's again likely that your expert checked this or knows the numbers, but it's not something you shared here. I'm curious whether the people who've bought your book feel differently from me.

And I've got no candidate, so I'll keep my 900 bucks. Just saying that as an interested impartial observer, from what I've read on this forum, I'm not convinced.

Derrick

Derrick, it is very easy to shoot from the rafters and just throw out generalities like 'bad math' or 'lie with statistics'. You also said it is easy to write a similar analysis software, but you can't back up your claim. I have put up an expert opinion in favor of Edward Gallaway, which says: "the common lexicon and the high number of four word phrases suggests a strong possibility of identity of authorship". For me that weighs heavily for Gallaway. At this point there is no other linguist expert opinion suggesting as strong an identity as this one for any other candidate. I am waiting for any other expert opinions on other candidates.

[DChung]

Derrick, it is very easy to shoot from the rafters and just throw out generalities like 'bad math' or 'lie with statistics'. You also said it is easy to write a similar analysis software, but you can't back up your claim. I have put up an expert opinion in favor of Edward Gallaway, which says: "the common lexicon and the high number of four word phrases suggests a strong possibility of identity of authorship". For me that weighs heavily for Gallaway. At this point there is no other linguist expert opinion suggesting as strong an identity as this one for any other candidate. I am waiting for any other expert opinions on other candidates.

No, I said it's easy to write a common substring algorithm. And the link I gave is code that does pretty much that, which took 2 minutes to find on google. That sounds pretty damn easy to me.

I don't claim to be able to do anything beyond that. In fact, my point was you NEVER told us what other analysis was done. Again, all I can do is speak about the evidence you've shared here, which is weak and sloppy.

As for bad math, I've laid out reasons why the evidence you provided aren't convincing and given you a reference to better understand the mistakes that people often make when dealing with statistics. Note that the title of the book was just for marketing sake. It really is a good introduction and shows various ways that data can be misinterpreted either deliberately or not.

And has anybody else gone to a linguistic expert to look for such an opinion? By your own account, the answer is no. I thank you for sharing this new tidbit, but what does "strong" possibility mean especially in the absence of having done this analysis with other candidates? Does that mean that 1% of writers write this way, or that that he puts the odds of him being the one at 10% or 25%.

In any case, you don't get to win the argument just because other people haven't run the test yet.
Bolt is the fastest runner in the world because nobody raced with him. He's the fastest because they did and they lost.

When and if the other candidates are tested and come out as worse matches than yours, then there might something stronger to say. Until then, all you got right now is this vague statement of "a strong possibility" which certainly isn't the same as an emphatic "That's the guy."

Derrick

[lybrary]

I don't claim to be able to do anything beyond that.

That is the general problem with your argument. Once you are able to do more I am happy to continue the discussion.

[Bill Mullins]

Bill, you want to reduce the knowledge of an expert to a number. That will not work.

With all due respect, you are who has reduced it to a number -- 130. What I want is a set of numbers, an accepted process to get to them, and comparisons to other numbers that show how the Erdnase/Gallaway numbers are of significance.

I appreciate that you've named your expert. Now I can read up on some of his previous work to understand how he comes to his conclusions.


BTW, you said: "For me the linguistic fingerprint is the strongest evidence one can present absent of any documentary evidence."

John Olsson, your expert, said: "Nobody has yet demonstrated the existence of such a thing as a linguistic fingerprint; how then can people write about it in this unexamined, regurgitated way, as though it were an established fact of forensic life?" [John Olsson, Forensic Linguistics: Second Edition. London, New York: Continuum International Publishing Group, 2008, p. 26]

[lybrary]

Bill, you want to reduce the knowledge of an expert to a number. That will not work.

With all due respect, you are who has reduced it to a number -- 130. What I want is a set of numbers, an accepted process to get to them, and comparisons to other numbers that show how the Erdnase/Gallaway numbers are of significance.

I appreciate that you've named your expert. Now I can read up on some of his previous work to understand how he comes to his conclusions.

Bill, I offered the data because I was hoping it would be of interest. Apparently it was not interesting and some got confused by the data. Why is nobody else willing to offer a similar analysis for some other book or candidate?

[mam]

If anyone wants any word sequence comparison between two texts, I already wrote a script for that. It looks at n-grams in a selected range (e.g. four words or more) in both texts. The default is to skip results with too many stop words (such as "of", "a", and other very common and generic ones) but it does not have to. I'll put the script on GitHub when I'm not on such a shaky connection. But in any case, just send me the text files and I'll do the comparison. This is by no means a proper linguistic analysis, I wrote it just to see if I could find any common relatively unique sentences in both Erdnase and Gallaway. The results from that is in this thread a few pages back.

[DChung]

I don't claim to be able to do anything beyond that.

That is the general problem with your argument. Once you are able to do more I am happy to continue the discussion.

I don't think you understand my argument at all then. So let me reiterate one last time. The evidence you've presented here is weak, and certainly not supported by the data (at least not from what's been posted here. That's it. Not sure how my ability to a forensic linguistic analysis has any effect on the strength of the data that you presented. Again, I'm just talking about what you've given here, not what you have stowed away or in your $30 ebook.

But as you haven't properly addressed a single one of the issues I've brought up regarding the problems with your mathematical analysis, then I'm not sure what else we have to talk about. Moreover you completely sidestepped my question about what "strongly possible" means.

It now seems to me that you're far more interested in pushing your candidate than actually discussing the merits of the evidence, in which case further discussion is probably pointless anyway.

Derrick

P.S. I don't think anybody was confused by the data so much as your analysis of it. We all get that there are lots of four-word matches. It's just that you haven't given enough context for us to decide whether the two texts are actually similar or that perhaps any two instructional texts are equally similar for example. Again, perhaps your linguist has done such an analysis, but it hasn't been presented here, and I've certainly got no reason to just take your word for it.

[lybrary]

P.S. I don't think anybody was confused by the data so much as your analysis of it. We all get that there are lots of four-word matches. It's just that you haven't given enough context for us to decide whether the two texts are actually similar or that perhaps any two instructional texts are equally similar for example. Again, perhaps your linguist has done such an analysis, but it hasn't been presented here, and I've certainly got no reason to just take your word for it.

Reading the books in question would be a good start. Apparently you haven't even done that. Dr. Olsson has. I am sure that factored into his expert opinion.