(11-10-2023, 07:48 PM)MarcoP wrote: You are not allowed to view links. Register or Login to view.I am not sure I understand correctly, but I don't know how these ideas explain the available evidence.
1. Though average word length is comparable for Voynichese and ordinary European languages, the shape of the distribution is significantly different. In particular, Voynichese has fewer short and long word tokens than ordinary languages (see You are not allowed to view links. Register or Login to view.).
2. A simple example of a many-to-many cipher would be a Caesar cipher where words are written in "Boustrophedon" style, i.e. alternating left-to-right and right-to-left for each line, so that 'wolf' can be ciphered as 'XPMG' or 'GMPX', and both 'XPMG' and 'GMPX' can be decoded as both 'wolf' and 'flow'. This would increase character conditional entropy (while we are looking for methods that decrease entropy) since each bigram would occur in both its original and reversed form with the same frequency (e.g. TH vs HT, CK vs KC). Also, this cipher does not explain why consecutive words tend to be similar in Voynichese (e.g. <f108r.42> okeey.qokeey.qokeedy.qokeey). Maybe this method is not what you are thinking of, but it seems compatible with the description you gave.
Here is an example from my text on distance based ciphers (one example of one-to-many almost stateless ciphers - with a local state that spans several character pairs at most - dictated by necessity to pick ascender/descender types that have no unclosed pairs at a given locus). At the top the word "mouse" is encoded in two different ways showcasing the decoding step by step, at the bottom there is a simple English phrase which demonstrates how all visible structure of plaintext is obliterated by this cipher. As you can see, there is no direct connection between the source word length and the target word length, and the entropy of the ciphertext is quite low.
Link to the section that explains how this cipher works: You are not allowed to view links. Register or Login to view.
I claim that after about an hour or two of learning and practicing this cipher it's possible to read it off the page without any external tools of reference tables and write it without using any special tools either. It is possible to improve this, by adjusting the cipher to make various kinds of mnemonic connections between the plain text letter shapes and the ciphertext (e.g., assigning wide distances for naturally wide letters, assigning first or second circular/crossbar/fork bases to characters of corresponding shapes).
Note that while all separate design principles of this cipher are mentioned at least in 1624 Cryptomenytices, and some of them in Traicté des chiffres ou Secrètes 1586, this specific mix is of my own design and I haven't seen this combination in literature. However, I suppose anyone who spent substantial time trying to work with distance-based ciphers would invent a similar scheme just by trying to gradually optimize the process both in terms of space that the ciphertext takes on a page and in terms of complexity.
I have no specific evidence at the moment, that the Voynich manuscript uses a cipher like this, but I have no evidence to the contrary either.