I thought that my idea of a lipogrammatic text without any of the letters ABC was about as extreme as a lipogram could possibly get, but now another idea has occurred to me along the same lines, and it may suggest a lipogrammatic text even more extreme than that:
If the ms is a product of an adherent of the English House of York, whose fierce rival was the House of Lancaster, then a lipogram excluding letters from the start of the alphabet such as A and C might indeed make sense. All the more so because York begins with a letter near the very end of the alphabet.
The more extreme version would be a lipogram that excludes all of the letters LANC from the text! Now this would invalidate the whole [qokedy qokeedy] = "nece neece" idea, but it would not necessarily affect the [otol] = "fous" idea. It would also explain even more of the problems we have recognizing words and grammatical endings of words, since the exclusion of "N" would change the appearance of many grammatical suffixes and other word endings significantly. It would also exclude all normal means of expressing negation in English!
To be sure, my provisional Voynich character : Middle English letter correspondence table would need to be amended in the case of a "no LANC" lipogrammatic text. But again, [otol] = "fous" and in fact all of the Middle English function words I proposed so far in the "problematic...Germanic" thread would not necessarily be affected by this adjustment. The only other word in my current hypothesis that would need to be changed is [qokaiin] / [qokain] = "no".
In general, the exclusion of more letters makes it more plausible to assign all remaining letters to Voynich characters or character combinations.
(17-04-2021, 06:22 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.The more extreme version would be a lipogram that excludes all of the letters LANC from the text!
With such extreme constraints, there would be little wiggle-room for more constraints, but we actually see the opposite in the VMS: vord statistics and glyph mono- and bigrams statistics show an evolution and a lot of variation in frequency that stand out at the quire, page, sometimes even at the paragraph level. Some have surprisingly high frequencies, others surprisingly low. Here are some lipograms: You are not allowed to view links.
Register or
Login to view. and You are not allowed to view links.
Register or
Login to view. don't have any EVA-e.
This (0 frequency on a page) happens so often on bigrams that I had this idea that pages could be "lipogrammatically linked" to the next page by the absence of some common bigram.
For example:
You are not allowed to view links.
Register or
Login to view. : 0 al
You are not allowed to view links.
Register or
Login to view. : 0 al
You are not allowed to view links.
Register or
Login to view. : 0 dy
You are not allowed to view links.
Register or
Login to view. : 0 dy
(17-04-2021, 08:29 PM)nablator Wrote: You are not allowed to view links. Register or Login to view. (17-04-2021, 06:22 PM)geoffreycaveney Wrote: You are not allowed to view links. Register or Login to view.The more extreme version would be a lipogram that excludes all of the letters LANC from the text!
With such extreme constraints, there would be little wiggle-room for more constraints, but we actually see the opposite in the VMS: vord statistics and glyph mono- and bigrams statistics show an evolution and a lot of variation in frequency that stand out at the quire, page, sometimes even at the paragraph level. Some have surprisingly high frequencies, others surprisingly low. Here are some lipograms: You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. don't have any EVA-e.
These pages without EVA [e] = Middle English "e" could represent an even more extreme version of this "Yorkist lipogram" that also excluded the "E" which is in "Lancaster" but not in "York".
nablator Wrote:This (0 frequency on a page) happens so often on bigrams that I had this idea that pages could be "lipogrammatically linked" to the next page by the absence of some common bigram.
For example:
You are not allowed to view links. Register or Login to view. : 0 al
You are not allowed to view links. Register or Login to view. : 0 al
You are not allowed to view links. Register or Login to view. : 0 dy
You are not allowed to view links. Register or Login to view. : 0 dy
Again, EVA [l] & [d] = Middle English "s", a letter which is in "Lancaster" but not in "York".
Just a hypothesis at this stage of course.
As a further refinement of this "Yorkist lipogram" idea, the general rule may only be "no LA" everywhere, but other letters in "LANCASTER" are selectively excluded from certain pages and sections. This would explain the examples nablator provided above, as well as the absence of EVA [qok] = Middle English "n" on some pages.
Yes, I am aware that other letters not in "Lancaster", such as EVA [qot] = Middle English "m" (?), are absent from certain pages as well. I'm not saying my idea can explain everything yet. But I think I'm on the right track. (Also, "m" is just a provisional hypothesis, has not appeared in key words like "fous" and "irour thee" yet, and maybe [qot] actually represents a different Middle English letter, possibly even "c" or "t" [not part of "th" in this system] to stick to the "Yorkist" theme of excluding "Lancaster" letters?)
Another observation: EVA [ch] = Middle English "y", EVA [o] = Middle English "ou", EVA [i] which is part of Middle English "o", and EVA [r] = Middle English "r" do not seem to be absent from any pages, except for You are not allowed to view links.
Register or
Login to view. which has almost no text on it at all. So perhaps the rule is that the letters of "York" are essentially present on every page, but other letters are not? (Yes, the apparent lack of any final "k" in this system is another issue that I do not have a "Yorkist" hypothesis to explain yet.)
Again, I understand that this is provisional and clearly there remains much that still needs to be figured out. But I think I'm on the right track.
Goffry
(17-04-2021, 08:29 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.This (0 frequency on a page) happens so often on bigrams that I had this idea that pages could be "lipogrammatically linked" to the next page by the absence of some common bigram.
For example:
You are not allowed to view links. Register or Login to view. : 0 al
You are not allowed to view links. Register or Login to view. : 0 al
You are not allowed to view links. Register or Login to view. : 0 dy
You are not allowed to view links. Register or Login to view. : 0 dy
Hi Nablator,
my statistical competence is limited, so it could be that I am misunderstanding something.
I have computed an histogram of 'al' counts per page on HerbalA pages (each bar counts the pages with that many occurrences). In my opinion, if we were facing the deliberate omission of 'al' on some pages, we would see two groups of pages: some with no occurrences and others with several occurrences. But here the most common situation is 1 occurrence and there is a considerable and apparently uniform spread. The number of pages with 2 occurrences is close to the number of pages with 0 occurrences. It seems to me that the simplest explanation is that the number of occurrences varies smoothly, with no break between "no-al" and "with-al" pages.
Possibly, looking at percentages instead of actual counts could tell us something more, but I don't have the time to do that right now.
(18-04-2021, 08:57 AM)MarcoP Wrote: You are not allowed to view links. Register or Login to view.Hi Nablator,
my statistical competence is limited, so it could be that I am misunderstanding something.
Hi MarcoP,
Some day I'll find the time to do a proper statistical analysis.
Well, if undesratnd correclty, with lipogrammatic technique the author would avoid, and hence omit, all words containing a certain letter or letters. It is not that he would simply omit undesirable letters in words. For example, avoiding the letter "a" would not make "pple" of "apple", it would just lead to omission of the whole word "apple".
Right, so all it does is reduce the alphabet. It doesn't mess with too many other stats. There may be collateral damage in some cases, like if you forbid "u", then "q" will also disappear. But your text will still look normal.
Unless if you do something extreme like omitting several vowels. For example if you can only use words with "o" or "u", your text will look very weird. This is impossible to do for a long text like the VM.
(18-04-2021, 11:33 AM)Koen G Wrote: You are not allowed to view links. Register or Login to view.But your text will still look normal.
Some things may be altered, e.g. if in English we exclude the letter "e", then, a) the character relative frequency table will be altered (frequences measured relatively to the most frequent letter will change), b) bigram frequency table will change as well, notably there will be no "e_" (e plus space) no more. Grammar will be more restricted - e.g. insofar there would be no "were", plural past tense will diminish in occurrence, etc.
Certainly, especially with a frequent vowel like [e], English would take a serious hit. What I mean is that if we'd look at the result as a new language (so not specifically spot the differences with normal English), any stat you can think of would be well within the range of normal linguistic behavior.