[Article] Harvard physicist's take on the VMS

[Rafal]

[*]

[*]Voynichese does not look like a straightforward representation of natural language --> agreed
Moreover, it cannot be readily understood as the result of common medieval cipher methods --> agreed
We don't know of anything like this --> agreed
ERGO, Voynichese must be the result of a nonsense-generating algorithm performed by the medieval scribe(s) --> what?

I feel there is more argumentation in the text.
[*]There is a chapter 5.1 Word order and vocabulary which tries to show that VM has a very weak word structure, different from real text. You cannot make a decent prediction a word based on the previous word, unlike in real languages.

[*]I am unable to evaluate the strength and validity of the argumentation, however.