Something that comes to mind is the Illiad in its Greek verse form. There would be quite some repetition for mnemonic purpose. Is that something you can test?
Alternatively, as has been suggested, a practical text that contains lots of instructions or directions might yield different results.
Quote:Something that comes to mind is the Illiad in its Greek verse form.
In my paper from 2014 I have used the The Aeneid from Virgil as latin text [You are not allowed to view links.
Register or
Login to view.]. The text class is the same as in the Ilias. In fact the Aeneid is Virgil's answer to Homer.
With You are not allowed to view links.
Register or
Login to view. I get the following graphs:
[
attachment=1120]
A smaller part of this graph also reveals multiple smaller networks with two, three and up to 30 word types:
[
attachment=1121]
Here is the gephi project:
[
attachment=1122]
(05-02-2017, 12:18 AM)Torsten Wrote: You are not allowed to view links. Register or Login to view.Quote:All you can possible say, at the very most, is that the Voynich manuscript is not written in Arabic, English, Latin, or German.
I used text from different text classes like a poem and verses. If you know a text class which is more repetitive then a religious text then indeed you can proof me wrong.
Why can't you even acknowledge that there's a problem? Why can't you even admit that you need to test more languages? Why can't you understand that you have to show there is a language-like pattern for co-occurrence before you can assert that the Voynich text doesn't fit it?
Torsten, you've made some great and interesting observations, but your theory is a leap into speculation and not remotely based on a firm argument.
Torsten could you please compare Chaucer's Canterbury Tales to see if it has a relationship with the VMS?
Quote:Why can't you even acknowledge that there's a problem? Why can't you even admit that you need to test more languages? Why can't you understand that you have to show there is a language-like pattern for co-occurrence before you can assert that the Voynich text doesn't fit it?
The co-occurence pattern is a feature of the Voynich text. If such a pattern did not exist for languages then maybe your assumption that the VMS must contain language is wrong. Your blog posts show that you know this already [see You are not allowed to view links.
Register or
Login to view.].
It is an error in my eyes if you compare languages to the VMS. You have to do it the other way around. If you want to decipher an unknown script the first question you should ask yourself is what features did characterize the script most. Only if you found such a feature you can start to search this feature somewhere else.
What happens if you interpret something into the VMS is illustrated by the attempt of Prof. Stephen Bax and by the attempt of Prof. Gordon Rugg.
Stephen Bax describes his approach as 'bottom-up method' following successful decodings of the Egyptian hieroglyphs and Cretan Linear B script. Bax describes this way as finding individual proper names in the text of the VMS. But he is misinformed about decipherments of the past. Finding proper names was in all cases only the last step. For instance in the case of Linear B it was Alice Kobers who noticed that words have common roots and suffixes. This way she was able to determine Linear B as syllabic script and to describe suffix changes for different roots. Only because Ventris had a basic understanding of the way Linear B was working he was able to decipher Linear B. It is simply wrong to argue that Ventris has deciphered Linear B by guessing some names. In fact he was able to read as 'ko-no-so' since he was knowing that the signs for /ko/, /no/ and /so/ are related to each other and share the same vowel [see You are not allowed to view links. Register or Login to view.].
In a similar way Champollion has identified the hieroglyphic script on the Rosetta stone as being written in a mixture of ideograms and phonetic signs and was knowing that the script doesn't write vowels before he started to read names [see You are not allowed to view links. Register or Login to view.]. It seems as Bax made the references to the decipherments of Linear B and of the Egyptian hieroglyphs without checking the details.
You can check each decipherment of the past the first step is always a deeper analysis of the unknown script.
Bax tries to guess 9 out of 8026 words in the VMS [see You are not allowed to view links. Register or Login to view.]. Bax admits in his methodic part that he didn’t know the language of the VMS, that he didn’t understand the way the script is working, that he is guessing the names for his nine words and that he is guessing the sound values for the glyphs. In a similar way Gordon Rugg admits in his paper that he is trying to simulate some of the features of the VMS as side effect of the Cardan Grille. Then Rugg tries to explain all the features of the VMS as effects of his Cardan Grille method. If you search for features found somewhere else in the Voynich manuscript you are lost before you even start.
The only way to decipher a script is to analyze the script. Only if you found a characteristic feature you can try to find it somewhere else. Because of this reason I have checked the rare words occurring only seven or eight time in the manuscript if I can see a pattern. The pattern I found was that this rare words occur near to similar words which are also rare [You are not allowed to view links. Register or Login to view.]. Since this feature is characteristic for the VMS I was able to do the next step. This step was to check for connections between the most frequently used word types of the VMS. By grouping the word types according to their similarity and frequency I was able to build the grid which is part of my paper from 2014. By building this grid I was building the network for the VMS step by step.
If you want to check a decipherment you should check the way the decipherment was done before you start to check the result. It is simply not possible to guess the correct solution for an unknown script.
That's a lot of words to say 'no'.
(05-02-2017, 12:46 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.That's a lot of words to say 'no'.
Even if there are a lot of words they contain phrases like "It is simply" or "to check". Or to say it in your words: "Even though the letters inside words are highly ordered, the words inside sentences seem to be wholly disordered. I do not understand how a language could do that." [You are not allowed to view links.
Register or
Login to view.]
If you take the three (EVA) letters 'e', 'o', 'l', and count the occurrences of their six permutations (in the Takahashi transcription), I think you get the following:
eol - 961
ole - 39
oel - 0
leo - 2
loe - 0
elo - 0
These numbers strongly imply that Voynichese has very strong letter adjacency rules in play: and it is surely far from coincidental that both eol and ole are the only two permutations where l immediately follows o. (For reference, 'ol' occurs 5507 times in the same transcription.)
Without any doubt, these numbers are inconsistent with auto-copying if you are trying to argue that errors introduced during autocopying are the systematic source of the random variation within Voynichese.
Thanks Nick, but don't hold your breath that Torsten will reply constructively.
I pointed out the problem of word structure to to him in 2014 after his first paper. Over two years later he still hasn't adequately responded to it, saying that the writer was free to write as he pleased within his 'concept of language'. I don't think he has an answer because it would spoil his theory.
It's sad, but he's another Rugg with a neat answer which is all wrong.
Nick , you do not even have to bother to compare combinations of a 3-letter word.
Just look at .ar. and .al. for example.
They have a repeat distance between each other which is steady throughout almost all "10 sections"
I can show the data but this only blurs the point here.
Always .ar. has the highest repeat and then follows .al. with also high but always lower repeat.
(to be exact, in the zodiac section ar and al seem the have the same or a slight reversed amount: al:29 and ar:27 depending on the transcript)
This fact alone already shows one of these two things:
1* the text is copied throughout the sections and therefore the text in all sections is the same.
or
2* the text is not copied throughout the sections but is "logically cohesive"
Clearly it is not 1), so it must be 2).