The Voynich Ninja

Yeah, I just can't get VMS out of my head… That's what I was afraid of.

The latest findings concern Middle High German / Bavarian.

The spacing pattern:

The transition from y to q occurs only 13 times within a word, but 3,635 times across a space. This fits a Germanic pattern: If -y encodes a verb ending and qo- absorbs the following prepositional phrase, then [...]dy [space] qo is simply standard Middle High German recipe syntax: a verb followed by a prepositional object.

The most common EVA endings are -dy (17.9%) and -ey (10.5%)—together accounting for 28.4% of all words.

In Middle High German, words with verb endings (-en, -er, -nd, -et, -it) make up 26.3% of all words.

28.4% versus 26.3%—almost identical.

This suggests that -dy and -ey encode Middle High German verb forms, not nominal inflections. The space then marks the boundary between a verb and its prepositional object—precisely the dominant structure in medieval German recipe texts! 

This is consistent with earlier findings: As previously shown, the frequency of the prefix y- corresponds almost exactly to the frequency of German verbal prefixes (ge-, be-, er-). Now the suffix pattern also points in the same direction.

We have now measured frequency matches across several independent levels—prefixes, suffixes, spacing patterns, word lengths, and article/preposition usage—all of which point to the same language. At a certain point, the accumulation of statistical matches makes a random explanation unlikely.
The VMS behaves statistically like a language—specifically, like Middle High German/bavarian

(19-03-2026, 08:04 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.tchodaiinoeody ???

You haven't answered my question: in which transcription did you find this word?

(26-03-2026, 04:32 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.

(19-03-2026, 08:04 PM)Ruby Novacna Wrote: You are not allowed to view links. Register or Login to view.tchodaiinoeody ???

You haven't answered my question: in which transcription did you find this word?

It's a mis-transcription of this on f51r, line 11: 

[attachment=14903]

"tchodaiin ?pheody"

Edit: unless you're asking which transcription file was used... if so my bad

@ Ruby, oh okay. Now I get what you meant by that question.

eggyk is right, You are not allowed to view links. Register or Login to view. 10—to be honest, I don’t remember which transcription that was. I got it from voynich.nu. I cleaned it up, not with AI, but back then using Notepad++. It’s been a while, but it looks like that was a mistake in there. Sorry...


i should better use AI to get a clean Version 

(26-03-2026, 08:10 AM)JoJo_Jost Wrote: You are not allowed to view links. Register or Login to view.Yeah, I just can't get VMS out of my head… That's what I was afraid of.

The latest findings concern Middle High German / Bavarian.

The spacing pattern:

The transition from y to q occurs only 13 times within a word, but 3,635 times across a space. This fits a Germanic pattern: If -y encodes a verb ending and qo- absorbs the following prepositional phrase, then [...]dy [space] qo is simply standard Middle High German recipe syntax: a verb followed by a prepositional object.

The most common EVA endings are -dy (17.9%) and -ey (10.5%)—together accounting for 28.4% of all words.

In Middle High German, words with verb endings (-en, -er, -nd, -et, -it) make up 26.3% of all words.

28.4% versus 26.3%—almost identical.

This suggests that -dy and -ey encode Middle High German verb forms, not nominal inflections. The space then marks the boundary between a verb and its prepositional object—precisely the dominant structure in medieval German recipe texts! 

This is consistent with earlier findings: As previously shown, the frequency of the prefix y- corresponds almost exactly to the frequency of German verbal prefixes (ge-, be-, er-). Now the suffix pattern also points in the same direction.

We have now measured frequency matches across several independent levels—prefixes, suffixes, spacing patterns, word lengths, and article/preposition usage—all of which point to the same language. At a certain point, the accumulation of statistical matches makes a random explanation unlikely.
The VMS behaves statistically like a language—specifically, like Middle High German/bavarian

This a bold statement I hope you decode it and it's reproducible.  To me that would be a miracle and language experts would be in awe.  Any ideas what daiin might mean from Middle High German/bavarian and if I'm late to the game sorry.

Before attempting to decode it, I recommend building out a model of how this encoding of Middle High German/Bavarian behaves when it encrypts known Middle High German/Bavarian texts of various types. You can then compare the outputs of your notional cipher to Voynichese. That iterative approach would help test the idea and highlight especially large gaps between the two types of text.

@Magnesium and oeesordy

I’ve already done that; there were some striking similarities to Voynichese, but also some areas where there was absolutely no match. I took a recipe text from the Breslauer manuscript and ran it against most of the statistical data available so far—a colleague here on the board had given me a tool that calculates these statistics.

But the problem is a different one: The cipher isn’t that far along yet. Parts of the language haven’t been deciphered at all. I’m still not at all 
sure about some parts of the cipher. Although “sure” is probably the wrong word anyway. 

I’m still searching for many details of the cipher; it’s a tedious process. Precisely because the language underlying the text isn’t clear yet—that is, what kind of dialect it is.

If you take the word falzen from 116, for example, there are many variants, some of which are very different, especially when written phonetically, as was customary at the time:

valten
vulten
volten
valtzen
faltzen
pfaltzen
falsen
pfalsen
faltz
pfaltz
vult

that are all variants of one word!!! different dialects...

If you have an absorption cipher that is also variable—in other words, a really poor cipher—it becomes, as I’ve written here several times, virtually impossible to decode.

And if it were also a strong dialect that nobody knows anymore today.... 

It’s a puzzle with nearly 30,000 pieces, but one where the image has faded over the centuries.

And that would be a very clear explanation for why no one has managed to do it yet.

Also because, strangely enough, most cryptologists assume that ciphers are logical and structured... which is, of course, nonsense, especially in the 15th century.

That would take years—that’s exactly what frustrates me so much.


But—and this is the crazy part—on the other hand, there’s actually a lot of evidence suggesting it could be MHD/Bairish. Even statistically speaking. And even Stolfi, who’s usually so skeptical, agrees, saying: “If it’s a European language, then it’s Bairish...


------
PS. On top of that, even I, as a German, naturally have trouble understanding texts written in a 15th-century dialect. There are countless words I don’t know, which I have to figure out on my own or even look up. How is one supposed to recognize these in a cipher?

@ oeesordy : Any ideas what daiin might mean from Middle High German/bavarian and if I'm late to the game sorry.

Right now, I'm interpreting “d” as sharp Bavarian “sibilants” = s / z / tz / ss
so
dain=
sein (bavarian = sain)

“sain/sein” = sein
“tzsain” =together, Stone, etc.
as a suffix in a word: “ssain/ssein” “zein/zain”

“ein” (Bavarian “ain”) is one of the most productive endings in German.
But I'm also wondering if “daiin” on its own doesn't have a similar meaning like “qo,” “o,” “y,” “dy,” etc...

I should add this:

It is interesting to note that with the code: d = sibilants, a = a / o, ii as multiple diphthongs or a single vowel (e.g., “u”), and n = n, we arrive at a frequency of approximately 3.5% in standard Middle High German recipe texts.

daiin+dain comes to: 2.57%

However, there are of course many possible variations of the code, so this isn’t a realistic comparison. So I’m still very much at the beginning here.... it is and remains difficult. But one must understand that daiin is not a single word, but a word plus an ending and possibly a middle part.

The German word “sein” (to be) appears in several other words and stands alone. But that would not explain the frequency by a long shot. That’s why there must be a cipher involved, but I can’t say yet exactly what it looks like.

----
As I mentioned earlier, this project is "work in progress". And the goal isn’t to directly decipher the text, but rather to get closer to a decipherment and to verify whether the hypothesis that VMS might be Bavarian has sufficient statistical evidence.

I can go ahead and publish the current status; this concerns words that have a certain significance (not all the possibilities found). And these have been compared with MHD recipe books (Ortloff of Bavaria, Breslauer, Bartholomäus, and a collection of cooking recipes). All of this, of course, with the help of Claude.

But this is just the beginning; I’ve only worked through a few passages so far because it’s very time-consuming.

(However, the data has not yet been cross-checked (this is a preliminary result), so it is subject to change.)

It's more meant to show how I go about finding good matches. Of course, I could also search for and write down all possible word combinations, but then I'd end up with hundreds of words. That approach, though, gets too bogged down in analysis.

The next step is to find more good matches and then identify the words surrounding those matches that are also good matches, in order to recognize more and more connections (if possible) an possible sentences.

[attachment=14910]

MHD total words: 173651