The Voynich Ninja

(28-06-2026, 11:08 PM)tavie Wrote: You are not allowed to view links. Register or Login to view.I'm not following what you mean by "substrate languages"

Dear Tavie, I used the term "substrate languages" to refer to the languages of the presumed precursor documents. For clarity,  I have chsnged the term to "precursor languages".

The "L-zero" glyphs and the Voynich alphabet:

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

The L-zero transliterations ... and the five most promising precursor languages

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

(28-06-2026, 07:24 PM)dfs346 Wrote: You are not allowed to view links. Register or Login to view."Targeting the precursor languages" on SubStack:
You are not allowed to view links. Register or Login to view.

Quote:My third assumption is the precursor languages are more likely than not to be languages that were written in Europe during the fifteenth century.  This assumption follows from the radiocarbon dating of samples of the Voynich vellum to the fourteenth or early fifteenth century; and from Wilfrid Voynich’s assertion that he had discovered the manuscript in southern Europe.

And that is where everybody got lost down the wrong path.  Sigh.

Those facts are undisputed, but that  assumption does not follow at all from them.  

The right approach would have been, "let's determine the properties of Voynichese, and then let's look for languages that have similar properties, modulo the original alphabet and spelling system."

All the best, --stolfi

more on alternative transliterations, the top twenty "words", and the Persian dimension ...

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

(31-07-2026, 10:05 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.And that is where everybody got lost down the wrong path.  Sigh.

Those facts are undisputed, but that  assumption does not follow at all from them.  

The right approach would have been, "let's determine the properties of Voynichese, and then let's look for languages that have similar properties, modulo the original alphabet and spelling system."
[..]

Right.
And was that ever done systematically for all european languages of ~1400AD? Did you?

(09-08-2026, 04:26 PM)dfs346 Wrote: You are not allowed to view links. Register or Login to view.more on alternative transliterations, the top twenty "words", and the Persian dimension ...

I think your efforts are wasted, and I suppose you know why I think so.

Apart from that, I see two problems with your approach: (1) you are treating the whole book as if it was a homogeneous text, and (2) you are comparing word frequencies with texts of the wrong type.

As for (1), even though the language seems to be basically the same throughout, the VMS is visibly at least four separate books on very different subjects: Herbal+Pharma, Astro+Cosmo+Zodiac, BioBalneo, and Starred Parags (Recipes).  Apart form the structure of the tokens, there is no evidence of any discourse thread spanning across sections.   Like many books of the time, each of these sections is probably a transcription or translation of a separate "ur-book", possibly by a different author and from a different time and place; each with its own style, emphasis, etc.  And Herbal may be a merge of two independent "ur-books", namely Herbal-A and Herbal-B.  

But then each section will have its "most common words" that are specific to that section.  Even basic function words like "and" and "is" can vary substantially in frequency between a herbal and a list of recipes, between a text on anatomy and a manual of astrology...

As for (2), the contents of each section of the VMS (except perhaps BioBalneo) is not going to be a discursive text, like the typical theology or philosophy treatise.  Its sentence structure is likely to be rather constrained and formulaic.  We don't expect nontrivial subordinate clauses or logical arguments.  Each section will have a specific set of frequent nouns, adjectives, verbs, and verb tenses.  Thus, when trying to match common words, you should try to use a reference text that is likely to have the same "technical" style.   

And even then you must be aware that two books on the same "technical" subject may have very different styles.   For instance, Culpeper's herbal tells many random stories about each plant, so it ends being similar to a discursive treatise; whereas the Alchemists Herbal is completely different, with very rigid and terse "formulas". 

All the best, --stolfi

(09-08-2026, 05:34 PM)Stefan Wirtz_2 Wrote: You are not allowed to view links. Register or Login to view.And was that ever done systematically for all european languages of ~1400AD? Did you?

"European" languages generally have polysyllabic words.  That is the case for all the Indo-European languages, for the Finno-Ugrian ones (Finnish, Sami, Hungarian), and for all the "outliers" I know about, like Basque, Turkish, and Georgian.  If one plots how many words there are with K letters, one gets a plot with a long tail, stretching well beyond 10-12 letters.  The same letters may appear several times in a word, with similar contexts.

The structure of Voynichese words does not seem to allow polysyllabic words. There is a rather sharp limit on the length of the words.  There may be at most one gallows glyph in each word.  The other glyphs cluster around the "core" of the word in layers, according to rather limited patterns.  That structure is consistent with many East Asian languages, but not with any of the "European" ones above.

Semitic languages (Arabic, Hebrew, and Ge'ez) generally have words with 3±1 syllables.   Some people have tried to match the Voynichese word structure that of those languages.  (Arabic script, in particular, uses different letter shapes at the beginning, middle, and end of a word; which suggests a way to get around the "one-gallows max" property of Voynichese.)  But AFAIK those attempts failed, because even 3±1 syllables  is too many.  

It is possible that Voynichese is an "European" language where each syllable is written as a separate word.  But I suspect that such a text would have the inverse problem: the "words" would be generally shorter and simpler than the Voynichese words...

All the best, --stolfi

(10-08-2026, 12:25 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view."European" languages generally have polysyllabic words.  That is the case for all the Indo-European languages, for the Finno-Ugrian ones (Finnish, Sami, Hungarian),[..]

I agree for Indo-European languages, Finnish and Hungarian - but I do not agree for most of the other finno-ugric languages --they have a clear tendency to short words with rather not so much syllables.

But we should discuss this no longer beneath this thread here;

on the other hand, I don't understand why this one is under "News" all the time.
It's a theory, and should be under "Th & S".

Mappings of the 11th through 20th ranked "words" in the Voynich manuscript, transliteration vb101.L0.4, to medieval Persian; and some English equivalents.

[attachment=17292]

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