(18-10-2023, 03:27 PM)oshfdk wrote: You are not allowed to view links. Register or Login to view.If the labels in the manuscript are indeed labels (each corresponding to some plaintext words or numbers), and sometimes they are located without any obvious sequence on the page (e.g., star labels), does this mean this method will have to restart at each label? What would be the correspondence between the encoded label length and the plaintext length?
Yes, a restart. I haven't studied the peculiarities of labelese, it could be a good angle.
Length is certainly a problem with short labels because the strong statistical constraints on all flavors of Voynichese have a cost: some shortening is expected after decryption if the plaintext doesn't have as many constraints (normal natural languages don't). I speculate that labels were often truncated, shortened as mnemonics. Another possibility in some cases is that two labels have to be concatenated to make a plaintext label. Look at the arrangement of central labels on the Rosettes page for example: 7+5 = 12.
Quote:If I understand the method correctly, the encoded label is between N and 2 * N characters long?
It depends. There could be a little compression when there are useful alignments in the table: several letters can be enciphered with one Voynichese glyph, or some cells of the table could contain more than one letter (usual Latin abbreviations like per/par, pro, us, um, qu...) or when n/m is abbreviated after a vowel: Sh could be special cases of (e, e) coordinates with a macron added for the omitted n/m letter. There is evidence that they used i with an attached macron (not ciphered) at the end of a few lines because of lack of space (written differently than n, not slanted) so they were familiar with a common Latin abbreviation and were likely to include some of them in the system.
![[+]](https://www.voynich.ninja/images/collapse_collapsed.png)
