Discussion of "A possible generating algorithm of the Voynich manuscript"

[ReneZ]

The hypotheses says "replacing one or more glyphs with similar ones" (Timm & Schinner 2020, p. 9). There is nothing said about small changes.

Torsten, you are quibbling about words, and you are contradicting yourself.

You write:

Line 25  [cheekey]     che->chee, p->k, che->e                              ED=3

This is not 'replacing glyphs with similar ones'. It is applying small changes (L distance 1) as you have been writing for years.

Whether certain words exist on any page is a question of the accuracy of any transliteration. No two are the same. The differences may be errors in either one. I know only too well that there are many in my own, but there are also many in the Takeshi transliteration.
As I wrote in an earlier post of today, the similarity between the two is 97.5% at the glyph level. The similarity in identifying word spaces is much lower.

I am sure that my figure of 22.7 of all changes requiring 3 or more 'steps' is correct, and for me this invalidates the hyptothesis. Others may judge differently.

Renegadehealer wrote:

Torsten Timm has consistently said that his autocopy method is one of the few practical pre-modern ways to generate large amounts of gibberish.

Are you sure it is practical? Have you looked at the details of the algorithm inside the app?
This algorithm is what the medieval scribe would have done.
I have not looked into it, but I understand that Nick Pelling has, and he commented that it has quite a number of very specific choices.

In my opinion this concept is modern and would not fit a medieval mind. The best a medieval person could do would be to throw a die, and this is making the process slow.