The Voynich Ninja

(03-10-2022, 08:29 PM)R. Sale Wrote: You are not allowed to view links. Register or Login to view.So, if the VMs 'languages' are correlated to the 'scribes', then some of the differences in the languages are potentially either due to spelling and/or orthography. Is that how things look??

Not really. If we had only two homogeneous Currier "languages" and no "dialects", variant spellings and subjects could maybe account for the large discrepancies, however the situation is much more complex. Many pages show abnormal-looking frequencies of some patterns. On small samples (paragraph and page) only large variations in frequencies are visible (small sample => low statistical significance) so we don't really know what is happening exactly at the page and paragraph level. At the section level we can be much more confident that the differences are not a statistical fluke. Some frequent patterns in some sections are rare or missing in other sections. There is nothing random or natural about, for example, avoiding almost all (EVA regex) "[eolr]o[aysd]" in Q13. The only exceptions are one "lod" on f77r, one "los" on f78v, and maybe "par,ody" on You are not allowed to view links. Register or Login to view. and "oteeo,dy" on You are not allowed to view links. Register or Login to view. if the small spaces are not significant.

Aga,

Not referring to the city - it's the heraldic insignia. It has two furs.

You are not allowed to view links. Register or Login to view.

(04-10-2022, 03:06 PM)nablator Wrote: You are not allowed to view links. Register or Login to view.On small samples (paragraph and page) only large variations in frequencies are visible (small sample => low statistical significance) so we don't really know what is happening exactly at the page and paragraph level. At the section level we can be much more confident that the differences are not a statistical fluke.

Hi, there, Nablator, you are right. I can explain what is going on. 
In my Slovenian transliteration/transcription (not fully adjusted yet), I noticed some words occurring several times on the same line and on several consecutive lines, like in the section below. If you know Serbian, you would understand what the words LEK and LEČITI means. It is the root for the words related to healing and healing remedies, pharmacy, and even religious psychological mystical healing. And since the vocabulary was not yet diversified, there was a lot of repetition.

The example of the two section is only partially adjusted, however I noticed that the sections with many labels also have lower entropy, because the listed individual words are then written in 'sentences' as they are used in different grammatical forms or for different meanings.

Pokedi.okaiw.kair.okal.sar.olkaw.oLEKawal.olrol.dl.
Pol.če.čol.tar.bom.okedi.dal.bug.oLEKeol.oLEKudi.okeol.
i.skej.Pokan.čeki.otaiw.olor.okan.PoLEKuj.Potar.oLEKan.
sol.čej.ralčEJ.bol.oLEKol.českej.dal.če kej.oLEKaw.oLEKUJr.
So.LEKej.lečedi.PoLEKaiv.dal.



Potol.šestči.oti.dar.šestej.sozdi.Pokudi.oLEKej.
Svol.čd.čedi.še.stej.oLEKi.dar.osva.LEKaiwo.Posvečedi.oraim.ozvoli.oroli.
sar.ol.olam.i.Pol.i.Poror.skedi.če.kedi.okam.budi.Pokol.čedi.olaw.
Pokudi.okuj.dar.olečedi.Pso.LEKUdi.rar.českej.otar.čedi.olr.stej.lor.
d.če.dej.Pokedi.ol.če.skej.olečdi.sar.or.i.kudi.če.tej.ran.budi.šekail.
ik.čedar.orarar.budi.Pokal.daiw.budi.oLEKedi.Pokedi.Poki.čedi.dam.


The word LEK is used throughout the VM text, but in some sections, it is much more frequent and more repetitive.

The repetition of suffixes, such as -am, -dam, -dy, -al (EVA aiin, daiin, dy, al) is related to the style of writing (first, second, third person).

This would surely cause low conditional entropy.

(27-01-2025, 06:03 PM)cvetkakocj@rogers.com Wrote: You are not allowed to view links. Register or Login to view.The repetition of suffixes, such as -am, -dam, -dy, -al (EVA aiin, daiin, dy, al) is related to the style of writing (first, second, third person).

This is very confusing. First/second/third person variations are based on grammatical rules.
What do you mean with 'style of writing'.

(27-01-2025, 06:03 PM)cvetkakocj@rogers.com Wrote: You are not allowed to view links. Register or Login to view.This would surely cause low conditional entropy.

No, definitely not. There may be some difference in conditional entropy, between highly inflected languages and less inflected langauages, but these differences are not significicant, and the conditional entropy of the VOynich MS text is way lower than that.

Please be assured that the Voynich MS low entropy is not explained by inflection, and it is incompatible with your approach.

(28-01-2025, 03:40 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.

(27-01-2025, 06:03 PM)cvetkakocj@rogers.com Wrote: You are not allowed to view links. Register or Login to view.The repetition of suffixes, such as -am, -dam, -dy, -al (EVA aiin, daiin, dy, al) is related to the style of writing (first, second, third person).

This is very confusing. First/second/third person variations are based on grammatical rules.
What do you mean with 'style of writing'.



By style of writing, I mean in most simple terms the way the verbs are used in the VM. Writing in the first person singular indicative mood is not that common for a longer text, but it is very useful in direct conversation. In English, this style could be recognized by the pronoun I and a verb in infinitive. In Slovenian, the pronoun is not used, since the person, gender, number and time is indicated by a suffix -am, spelled also as -im, -em.  When there is a string of vowels in a line (not connected with comma or conjunction), they would have the same inflectional suffix (like 27  EVA m/SLO-VA il suffix in f3v. Therefore, it is safe to assume that the text containing high frequency of these suffixes is written in the first person singular indicative mood. This style of writing or speaking is often used in prayers (like Act of Contrition) and in direct communication. 
I understand that EVA aiin is just a transliteration and that it it is counted as  3 different glyphs in EVA. However, as Dr. Bax pointed out, and as it is evident from the medieval 15th century manuscripts, the final three minims - iin can be transliterated as different combinations, in which case they form different letters.
Without knowing the language, the iin combination can be transliterated as m, in, iiv, iw, inv. Because the handwriting was such that there is very little distinction between the minims, the proper transcription can only be done manually by somebody knowing the language. We know that those are mostly the suffixes. Knowing which language has such variety of suffixes is very helpful. Slovenian does! Computer cannot make a distinction, even if there was a program to train AI to take into account grammatical rules, because those rules were not standardized yet.
By 'style of writing' I also mean various peculiarities of the VM, like dropped semivowels that makes text look like abjad, like improper spaces after r, ch, sh, d, which causes too many spaces, or the lumping of two or three short words together, termed by contemporary linguists as 'words block', which account for missing spaces.  

The assumption that the high frequency of EVA daiin and aiin represents 'Language A' , and the part where dy is unusually high, is 'Language B' it totally wrong, because daiin and aiin (transliterated with SLO-VA as dam and am) and dy (transliterated with EVA and SLO-VA as dy) are just two different inflectional forms of the verbs 'to give', although these suffixes can also be used in different words.  
Because the word dy (now day) contained a semivowel that was dropped, it is spelled in the VM as DY. In the early Slovenian writing, the semivowels were replaced with full vowels, so that the word DY became dai/dei/dii, and after y was changed, to daj/dej/dij). The vowel in still pronounced as semivowel in dialectal speech. 

I have to thank you for providing the EVA alphabet, because it enabled me to recognize VM language and improve the alphabet.
Because of the unclear handwriting, EVA transliteration needs corrections and not even with SLO-VA transliteration/transcription can proper transcription be made. Besides, the simple substitution only works for the letters that have the same sound in Latin and Slovenian, for the rest of the words, another transcription is needed to transform Latin letters to medieval Slovenian, and another to transform the text to contemporary Slovenian. If this is not complicated enough, consider four different writing convention (German, Italian, Hungarian, Croatian) various writers used in the 16th century for the same language. 
 
 

(27-01-2025, 06:03 PM)cvetkakocj@rogers.com Wrote: You are not allowed to view links. Register or Login to view.This would surely cause low conditional entropy.

No, definitely not. There may be some difference in conditional entropy, between highly inflected languages and less inflected langauages, but these differences are not significicant, and the conditional entropy of the Voynich MS text is way lower than that.
If the conditional entropy is low when the letter O can be predicted to follow P (like in Slovenian prefix used to form perfective verbs), or in -suffixes  -dam, or dy, transcribing them only as such eliminates other possible biglyphs and increases rigidity.

Please be assured that the Voynich MS low entropy is not explained by inflection, and it is incompatible with your approach.

The example of the word LEK (root word for healing, medicine)  pertains to the style of writing and inflection. While the word is scattered all over the VM text, it is concentrated in some section to such a degree, that there can be two or three in the line and even in consecutive lines. I encountered this issue in the section where labels follow text. While similar individual words are used as labels, they are used in the text for different meaning in homonyms or in different inflectional forms. 
This part of the VM looks like the lost Slovenian dictionary and grammar book, written about 1470 (or copied from the VM) allegedly by Thomas of Celje, who was a secretary of Emperor Frederick and taught Emperor's son Maximillian to speak Slovenian. The book was mentioned by Slovenian 16th century writers, but was later lost.

How can you assure me that the h2 in Voynich MS is not related to inflection? While it is true that inflection provides variety, wrong transliteration/transcription of the most frequent suffixes, like -dy and -am can also cause too much rigidity, so does too much repetition of the same suffixes.

Cvetka, you are always welcome to discuss your theory in its own thread, but please keep it out of other threads. If everybody went "well, in my theory..." on these subjects, the forum would no longer be able to function the way it does.

And why Slovenian? The logic you present could equally be applied to any Slavic language. Is your hypothesis that the manuscript is in vowel-less Slavic?

Please continue the discussion of Cvetka's theory here:
You are not allowed to view links. Register or Login to view.

I think it would be useful to also formalize "autocorrelation of word lengths", "similar words occur close to each other horizantaly and verticaly" and LAAFU effects (different character distributions in the start, center and end of the line).