05-04-2020, 10:39 PM
Background: Ever since I got into entropy (thanks Anton) I've had my eye on "minim clusters" like [ain] and [aiin].
Last year I You are not allowed to view links. Register or Login to view. with modifying EVA transcriptions in order to increase entropy. The most effective measure I was able to find is to replace things like [aiin] by a single character. For example, [dain] becaomes [d1], [daiin] becomes [d2]. This measure alone bridges 1/3 of the gap between an EVA transcription and a low-entropy normal text.
(Side note: combining different measures has unpredictable results, and after a while there are diminishing returns. Solving one entropy issue just shifts the problem to others).
Now, I know there are exceptions, but this thread is not about those. The general tendency is that EVA-[i] or [ii] is preceded by [a] and followed by [n] (and to a much lesser extent [r]).
In other words, if I say EVA-[ii], you can pretty safely add [a] and [n]. This is why entropy is low, because the glyphs are so predictable.
Now a few days ago, Lisa Fagin Davis You are not allowed to view links. Register or Login to view. that "When [a] is followed by a minim, it ligates to the top of the minim. When [a] is followed by just about anything else, it does not ligate. But until we know what language this represents, we can’t know if this rule is paleographic or linguistic."
All these examples are from You are not allowed to view links. Register or Login to view. :
The ligation looks rather variable, but that's not the point. What I wonder is: how can we know whether these are all variations of the glyph EVA calls [a]? Don't some of these look more like a bunch of minims with a loop on the left and right? Like an "m"? In the same way "anchiton" could also be "michiton"? If [aiin] really contains the separate unit [a], then why is it always there? Might what we call [a] preceding minims just be what the VM minim looks like when it's not preceded by a minim? Like an onset flourish? How can we know?
Last year I You are not allowed to view links. Register or Login to view. with modifying EVA transcriptions in order to increase entropy. The most effective measure I was able to find is to replace things like [aiin] by a single character. For example, [dain] becaomes [d1], [daiin] becomes [d2]. This measure alone bridges 1/3 of the gap between an EVA transcription and a low-entropy normal text.
(Side note: combining different measures has unpredictable results, and after a while there are diminishing returns. Solving one entropy issue just shifts the problem to others).
Now, I know there are exceptions, but this thread is not about those. The general tendency is that EVA-[i] or [ii] is preceded by [a] and followed by [n] (and to a much lesser extent [r]).
In other words, if I say EVA-[ii], you can pretty safely add [a] and [n]. This is why entropy is low, because the glyphs are so predictable.
Now a few days ago, Lisa Fagin Davis You are not allowed to view links. Register or Login to view. that "When [a] is followed by a minim, it ligates to the top of the minim. When [a] is followed by just about anything else, it does not ligate. But until we know what language this represents, we can’t know if this rule is paleographic or linguistic."
All these examples are from You are not allowed to view links. Register or Login to view. :
The ligation looks rather variable, but that's not the point. What I wonder is: how can we know whether these are all variations of the glyph EVA calls [a]? Don't some of these look more like a bunch of minims with a loop on the left and right? Like an "m"? In the same way "anchiton" could also be "michiton"? If [aiin] really contains the separate unit [a], then why is it always there? Might what we call [a] preceding minims just be what the VM minim looks like when it's not preceded by a minim? Like an onset flourish? How can we know?
![[+]](https://www.voynich.ninja/images/collapse_collapsed.png)