Of bookshelves and statistics

[DG97EEB]

I have been struggling a lot with the transliterarion of the VMS. You migjt remember some post of mine where I saw some slight long distance word relationship. After all this work, I guess there was a method to write the manuscript. I am not sure if it has any sense at all, but a method, a way of writting all the glyphs forming sort of words, there was. What puzzles me more is how come this is so complicated to know if it is gibberish or it has sense. How could this be consteucted at the XVth century? So this slight signal in Q20... will it really help if the rest of the VMS has no signal?

I think I've built more than 35 generators and some of them come much closer than yours in terms of hitting stats including long range and yet none of them matches on everything. There is always space underneath for a potential source, but if course both the source and the exact mechanism to create Voynich on top remain opaque. I think option A is genuinely generative and then we have 200+ pages of arbitrary nonsense. I don't think the data quite yet supports that. Option B,  it's source specific and the Voynich is some kind of harmonisation layer on top, with maybe an interim cypher layer, but if that were the case, I don't see how we can ever reverse engineer it. We are then forever looking for a needle in a haystack.. but your generator work is very good, and all of these ideas help us narrow the constraints I think..