Why and how the text could be Bavarian

[JoJo_Jost]

You’re misunderstanding something here. The qokedy loop was a feasibility study, NOT a translation. It explored whether dividing the VMS into vowel bridges and consonant clusters could produce natural speech—something I had previously proven—and whether this would result in a comparably low entropy. I had proven that as well. Then I also clarified whether such extreme repetitions could be read as normal language. And that worked, too. (But it was extremely time-consuming for the two forms I tested.)

But I still can’t translate any passages. And, as I’ve emphasized here several times, I unfortunately suspect that this won’t be possible either. Because there were too many dialects in German, and there were no spelling rules for them.

A cipher like the VBM is extremely dependent on the underlying text. Without any clues here, there are an absurdly high number of possibilities—so much so that any frequency comparison falls flat. And without a starting point—a crib—even AI fails at this—at least so far.

Example:

“Pfeffer” = pepper
Spellings:

“pfeffer, pheffer, pfeffir, phefer, pfefer”

“Salz” = salt
“salz, saltz, salcz, salcze, saltze”

“Und” = and
vnd, und, vnnd, unnd, vnde, unde, vnt, unt, vndt

Take the phrase “pepper and salt”—that alone yields 225 possibilities with just 3 words. And when you look at how different the consonants between the vowels are, it becomes clear pretty quickly that you have to know quite precisely which spelling and which dialect the language is written in to translate it back without a key.

That’s where the VBM will fail. I can prove that it’s a possible solution, but I’m no longer convinced I’ll be able to find the real key—unless I come up with a completely different idea to find something like a crib… and I’m working on that.  That’s why I’m examining the statistics very closely to see if I can find a solid anchor here. I’ve had a few before, but they didn’t hold up.

So I once created a counter-test, though I would consider it merely a “theoretical” test. I ran it through the AI to see how many consonant clusters one would need to know in order to translate the text coherently. The AI came up with about 60 percent (!!!). I think that says it all.

But maybe someday the AI will be advanced enough to translate the VMS. Then, of course, I’d be happy if the basic idea behind the VBM actually proved to be true - if I’m still around to see it... 


PS. Interestingly, the VBM would thus also provide a very clear answer as to why the VMS has not yet been deciphered...