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

[Koen G]

Very interesting, Marco.
It seems to me that Timm's method is intended to mimic exactly these statistics, and apparently he did a decent job.

What I've been thinking for a while already though, is that someone with enough the skills could probably make something similar for Latin. It would need much more freedom than the Voynichese generator, but you could easily tweak it to generate something  that's phonetically regular in Latin and approaches Latin's edit distance and TTR stats.

So my feeling is that these stats, while very illuminating, certainly don't tell the whole story.

Rene: you mean for example sort the words of both texts by frequency and substitute in that order?