Jeff wrote:Another explanation would be two scribes applying a non-deterministic algorithm (for example, "transliterate this Latin text into Voynich characters, add some of the designated null characters wherever you want, and if a character is 'final', be sure and leave a space after it.")Has anyone considered the idea that the pages may have been mixed up on purpose by the author? (I would point out that, even though the placement of nulls is non-deterministic, this process would be completely reversible to someone who knew Latin and could figure out where the real word endings were.) I envision two scribes working from a stack of plaintext pages, each applying the algorithm more or less consistently within his pages but making different choices from the other scribe as to frequency and selection of nulls. As each scribe finishes his page, he adds it to the pile which will become the VMS and starts on a new one, thus causing the interleaving of languages A and B when the encrypted pages are put in the same order as the plaintext ones. Bruce |