I think I need to clarify what I meant in my last post, because now that I reread it, it sounds as if I'm arguing that no progress is possible without a deciphering. What I meant to say is that if a method provides a sensible deciphering of a large part of the text in a reproducible manner, then it doesn't matter if it can't explain some of the features and peculiarities of the manuscript. In fact, nothing trumps a deciphering that is reproducible and successful by consensus, and it doesn't matter if LAAFU or word length distribution or anything else would turn up a hard to explain emergent property due to some complex interplay of the encoding scheme and the structure of the plaintext, that doesn't follow from the encoding scheme alone.
On the other hand, I think it is still very valuable to have a scheme that seems to align well with many known properties of Voynichese, even if no deciphering is achieved. However, specifically for the space vowels no statistical study has been done yet and I don't see yet now it could reproduce the statistics of the manuscript.
On the other hand, I think it is still very valuable to have a scheme that seems to align well with many known properties of Voynichese, even if no deciphering is achieved. However, specifically for the space vowels no statistical study has been done yet and I don't see yet now it could reproduce the statistics of the manuscript.