I just picked an easy example, but I could also have picked chedy and shedy, which differ by only a
curl above the first letter, whereas in the proposed Greek translation they are totally different:
Quote:1. chedy = *γέτος = ϝέτος, variant of ἔτος,
2. shedy = σχέδη (or Latin scheda),
This was also just to demonstrate the enormous freedom in selecting Greek equivalents of Voynich characters.
This was also very clearly demonstrated by @nablator:
Quote:q = δ
d = θ/τ/δ
s = σ/ς/ζ
a = α/αι
e = ε/ι/ω/αι
o = ο/α
ee = υ/αι
al = αλ/αι
ol = α/αλ/αι
y = ε/η/αι/ης/ος/ως
ch = γ/κ/ϝ
Sh = κ/σκ/σχ
k = ν/μ
n = ν
which is even more freedom than I expected.
To clarify: when I wrote: "not fundamentally different" from simple substitution, I still consider this to be "not fundamentally different". Individual cipher text letters are replaced by individual plain text letters, consontants (seem to) remain consonants and vowels vowels. There is a lot of freedom to play with.
That is clearly necessary in order to be able to match more words.
An alternative 'method' used by many people with a proposed solution, is to allow oneself to select words from several different languages. In this case: several different dialects.
Rather than translating Voynichese to Greek, in my opinion this is more like projecting Voynichese onto Greek.
I write all this without any animosity or negative feelings towards Ruby, of course. This is just my honest opinion, in response to the question whether it can be demonstrated that the Voynich MS text is not (ancient) Greek.
In my further opinion, the problem is not the language, but the method. Voynichese is so far removed from all European languages, that a very similar method would be required to match Voynichese to any of these.
In other words: as soon as anyone finds a method to bring Voynichese closer to Latin, the same method will also bring it closer to German, Italian, etc.