19-09-2020, 10:14 AM
If I were to develop this into a full theory, I'd first strive for more consistency, like always treat o as a modifier or null.
That's not what I'm after though. It's just an experiment driven to the extreme. You could keep on doing this until your character entropy matches word entropy. And even allowing for some ridiculous transformations, the entropy values remained absolutely bottom tier.
That said, I like to know this information. Entropy and too-low information content are often mentioned by those who criticise Voynichese as natural language. Knowing that h2 can be normalised while keeping h1 in check is good to know for a more nuanced discussion about these matters.
But the bottom line remains that entropy is not a measure of a language. It is a measure of the way a language is written. And for Voynichese, this value is extremely low.
That's not what I'm after though. It's just an experiment driven to the extreme. You could keep on doing this until your character entropy matches word entropy. And even allowing for some ridiculous transformations, the entropy values remained absolutely bottom tier.
That said, I like to know this information. Entropy and too-low information content are often mentioned by those who criticise Voynichese as natural language. Knowing that h2 can be normalised while keeping h1 in check is good to know for a more nuanced discussion about these matters.
But the bottom line remains that entropy is not a measure of a language. It is a measure of the way a language is written. And for Voynichese, this value is extremely low.


![[+]](https://www.voynich.ninja/images/collapse_collapsed.png)