I often wonder about Voynich A and Voynich B and why there even are two different 'languages'. It is likely not that important to most solutions but it always seemed strange to me that there even would be two different 'languages'.
Since I am personally in favour of a cipher solution, my favourite theory is that Voynich A had a 'flaw' in its cipher method or simply that Voynich B was a new and improved cipher. I'm also partial to the idea of the same cipher being used on two different underlying languages.
What is everyone else's personal theories?
(08-10-2025, 07:11 PM)Skoove Wrote: You are not allowed to view links. Register or Login to view.I often wonder about Voynich A and Voynich B and why there even are two different 'languages'. It is likely not that important to most solutions but it always seemed strange to me that there even would be two different 'languages'.
Since I am personally in favour of a cipher solution, my favourite theory is that Voynich A had a 'flaw' in its cipher method or simply that Voynich B was a new and improved cipher. I'm also partial to the idea of the same cipher being used on two different underlying languages.
What is everyone else's personal theories?
My favorite theory is still that Voynichese is not cipher, but an exotic language that the Author recorded, in a phonetic notation of his own invention, from dictation by a native speaker.
So my best three guesses for the A/B difference is that (1) the Author changed the spelling system between the two sets. or (2) the Author switched to another Dictator with a different dialect, or (3) the two sets come from two different source texts written in different dialects.
However, I have not looked at the A/B differences carefully enough to say whether they are compatible with either of these hypotheses.
Independently of my theory above, I think it is very unlikely that the text was composed by the Author himself. Given the range of subjects and the differences between the organization of the various sections, I think each section is likely to have been copied from a separate book that had been written by someone else, possibly abridging or summarizing it.
All the best, --jorge
(08-10-2025, 07:11 PM)Skoove Wrote: You are not allowed to view links. Register or Login to view.It is likely not that important to most solutions
To the contrary. Solutions that cannot explain or ignore the variations in 'language' statistics can probably be discarded just because of that.
A and B exist on a single page, separated by paragraph. I'd argue 2 connected pages.. but that's another thing.
It seems unlikely that someone would use a flawed cipher in conjunction with the new version on a page.
Imo, if it is a cipher, the cipher changed. Not it's use, understanding, words used, self citation, dialect or subject matter (being the things to blame).
Things that simply do not happen in A, happen in B. You can delete "edy" entirely and this is still true.. though I guess it more accurate to say doesn't happen for about 100 pages. It depends what you call A and B beyond ed/edy.
[edit] I guess, thinking about it, they changed it for some reason so it being faulty somehow is as good a reason as any. Q8 is the most interesting section for this "change", just a shame a bunch of it is missing... still seems odd to do a half and half page if you already considered your cipher faulty.
It's not just a simple matter of "A langauge" and "B langauge".
I am still of the opinion that the language clusters can be explained by there being gaps of time between the writings of the sections. Most probably the VMS was written in an alphabet that the authors invented for the hoax manuscript and invented a method for generating the text. Probably they used their method no-where else and through lack of regular use they gradually lost some fluency in it in between writings. I mentioned this in an earlier post.
You are not allowed to view links.
Register or
Login to view.
(09-10-2025, 02:05 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.It's not just a simple matter of "A langauge" and "B langauge".
What do you mean? Do you see other statistical discontinuities?
All the best, --jorge
I do indeed. The sense of discontinuity is coming from the herbal section, where the alternation of herbal-A and herbal-B texts is simply striking.
Intermediate forms exist, and so do variations inside the A and B texts.
I wrote about this here: You are not allowed to view links.
Register or
Login to view.
(09-10-2025, 11:43 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.I do indeed. The sense of discontinuity is coming from the herbal section, where the alternation of herbal-A and herbal-B texts is simply striking.
Intermediate forms exist, and so do variations inside the A and B texts.
I wrote about this here: You are not allowed to view links. Register or Login to view.
Thank you René, your RZ language identification is very interesting. You may know I am working with automatic topics at paragraph levels. What I think lately is that the topics found are more style of writting than really topics. I have copied your last table (if I shouldn't please let me know and I will delete it) and I have run my topic detections in 7 topics, just to see if we have the same results. Here are the images:
[
attachment=11611]
[
attachment=11612]
(In my plot, each topic is represented by a color, each folio is a vertical stroke and in the vertical stroke, each paragraph is represented by a color proportionally to the paragraph length in the folio).
I see some differences with your table (remember that my analysis is per paragraph). For example, for hand 1, my model detects that is using 4 topics and your say 3. I think that, as I work at paragraph level, hand 1 green topic could be dispersed if we work at folio level (see that the proportion of green is very small agains the orange topic), so we can say that your results and mine are simmilar for hand 1. So your dialect C should be topic blue, your dialect A should be topic pink and your dialect Ae should be topic orange. But note that according to my model the scriba uses also in some paragraphs topic green which could be dialect Ce (not clear).
If we look at hand 3, it has 6 over 7 topics represented (just as in your study), but I think in mine, hand 3 is using topìc orange (which may be your Ae dialect) instead the pink topic (your A dialect).
Well, I hope you find it interesting. If you want any information about my work, feel free to ask.
Thanks, this is quite interesting. Note that I have not tried to find the cause of the dialect differences (at least not in that page).
Understanding the details better may hopefully lead to finding a cause, or reason, for the differences.