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

[nablator]

The purpose of this exercise was not to create a model for the Voynich MS text. It was jut to show two things:

- the same text can have very different behaviour depending on how words are defined
- it is possible for a meaningful text to show a correlation between edit distance and vertical distance in the text

Actually this correlation that we have been busy replicating with meaningful texts needs more analysis. If I am not mistaken, when the vertical distances are restricted to the same page the diagrams are much different:

Herbal A:
   

(Only You are not allowed to view links. Register or Login to view. has enough lines for vertical distance as large as 24-25.)

Quire 20:
   

(Only You are not allowed to view links. Register or Login to view. and You are not allowed to view links. Register or Login to view. have enough lines for vertical distance as large as 51-53.)

Global:
   

(Labels removed in all calculations.)

So the strong correlation seen on figure 4 (page 8) (that includes long distances) does not actually extend to intermediate distances (15-50 lines), except across pages. Since large vertical distances are more likely to fall across two different pages than a short vertical distances, any difference between pages (a different setting/preference/whatever on each page) translates to a misleading gradual increase on figure 4. On my diagrams the local increase is more limited over a few lines (7 to 15) and then there is a large flat (or decreasing!) area until the end of the page where a spike is observed only on the largest pages.

It's weird. I really hope I'm wrong.

Edit: not so weird because of the very low number of samples (1 or 2) and shorter end-of-paragraph lines.