25-03-2025, 04:16 AM
(22-03-2025, 04:02 PM)nablator wrote: You are not allowed to view links. Register or Login to view.(22-03-2025, 12:19 AM)Bluetoes101 wrote: You are not allowed to view links. Register or Login to view.(21-03-2025, 10:00 PM)newamauta wrote: You are not allowed to view links. Register or Login to view.I'm curious, what do you think of her explanation for low entropy and the decoding framework she came up with (from the 39-50 min mark in the video)? Correct me if I'm wrong, would that still be considered a substitution cipher? If she's saying the symbols represent phonetic sounds and "fadas"/accent marks, and the grouping of the symbols changes their meaning?
I'd just note that she says "a, o" are vowels and "l" accents vowels by preceding them, and "l" isn't really a letter which caused all this entropy confusion.
l - 9957
la - 429
lo - 577
Thing is, even if the idea was right and we ended up with less letters, the entropy would be worse right?. 1000 letter alphabet, lots of possibilities. 2 letter alphabet, not so much.
Or is it adding accentuated characters?
I disagree that l alone, as she claims, explains the low (conditional?) entropy h2, it is a much broader phenomenon. The effect of what she proposes should be trivial to test if we knew the exact rules. Is her mapping of Voynichese glyphs to Irish letters visible somewhere?
Grouping frequent glyph patterns including l into a single (accentuated) character may bring a modest increase of h2, probably on the 3rd digit IIRC: this type of modification of the transliteration has been studied by René Z. and Koen G. among others.
Maybe there is a confusion with Shannon's entropy h1, that only takes into account character frequencies, as the formula shows.
Here are the glyph substitutions I grabbed from the podcast video