[split] Does Voynichese evolve from A to B?

[Jorge_Stolfi]

So you're agree that a gradient capturing the evolution could exist in between sections then? Something like the 'spelling reform' you describe?

IF the A/B difference is indeed a spelling reform, sure, it is probably the case that all A sections were drafted before the B sections, or vice-versa.

But my preferred guess (guess only, no evidence) is that Herbal-A and Herbal-B are transcriptions of two distinct herbals, and the differences are due to slight dialect changes, different tense, different emphasis, etc.

My lalamo says that the German spoken in the Frankfurt area differs from standard German in ways that resemble the A/B differences: "swallows word endings (e.g., haben becomes hawn), and changes the "ch" sound to a soft "sch" sound (making ich sound like isch)".  It also tells me that Norwegian and Danish are almost the same language, but Norwegian pronunciation is "clear, uses distinct pitch accents (tones), and consonants are pronounced crisply" whereas in Danish "soft consonants like d, g, and b often turn into semi-vowels or disappear entirely", and there is a phonemic glottal catch that is absent in Norwegian.

A style change that could produce significant digraph frequency shifts would be from conditional ("if someone ate this berry, he would see pink elephants") to descriptive ("eating this berry makes one see pink elephants") or imperative ("eat this berry and you will see pink elephants") or telegraphic ("eat berry, see pink elephants"), etc.

An emphasis change could be: Herbal A is mostly about tonics for preventing baldness and bad breath, Herbal-B is highly concerned with remedies for heartburn and ingrown toenails.

As you may know, I became convinced that daiin in the Starred Parags section is most often a keyword meaning roughly "mainly for" that precedes a list of diseases, while chedy is most often a certain invisible fluid that carries the nutrition value of food from the stomach and spleen to the lower torso and performs several other important physiological functions around the body.  If that is the case, obviously the frequency of use of those words (and hence of the digraphs they contain, like the infamous ed) is expected to be different in other sections.

And I also am fairly convinced that the endings -ir and -iir were actually -iin and -iiin, and that many k (but not t) were actually d, which were written by the Author in sloppy semi-cursive handwriting on his draft, and thus misread by the Scribe.  (Possibly those errors also affected many other -r endings, so that many dar may have been originally dain; but my word model allows dar while it excludes dair, so I am still quite uncertain about these.) If that is the case, the frequency of such replacements in other sections may be higher or lower, depending on how sloppy was the Author's handwriting on the respective drafts.

All the best, --stolfi