The Voynich Ninja

(17-06-2020, 12:32 AM)Torsten Wrote: You are not allowed to view links. Register or Login to view.
...First, it is necessary to analyze the way the writing system works. ...

YES!!!!! I agree absolutely 200%.

Hannig has stated this:

Quote:So bleibt als wahrscheinlichste Alternative nur die Variante a), also eine natürliche Sprache ohne Kodierung, aber in einer neuen Schrift.

For nonGerman-speakers, he is saying that  "the most likely alternative" is that Voynichese is a natural language, unencoded, but written in a new script.

The rest of the paper is based on this assumption.

Hannig's analysis

As a starting point, Hannig assumes that a o y e and i are vowels.

He then compares VMS "vowel" patterns to a number of languages, giving three languages as examples (English, German, French).

Based on vowel patterns and the shortness of the VMS tokens, he concludes that the language cannot be Indo-European, Basque, Turkish or Finno-Ugric, that it is a Semitic language consisting mainly of three consonants.

Based on this, together with European-style drawings, he writes that there are only three possible options: "Arabic, Hebrew or Aramaic".

Quote:Wegen des (durchgängigen) Fehlens des arabischen Artikels el/al ist als wahrscheinliche Lösung nur Hebräisch als Sprache des Voynich anzunehmen, da Aramäisch in Europa nur im Zusam- menhang mit dem Hebräischen genutzt wurde.

Thus he deduces that Hebrew is the only probable language.

Quote:Wenn man sich intensiv mit dem Voynich-Text beschäftigt, sucht man zwangsläufig nach einem Artikel. Die vielen Worte, die mit  anfangen, z.B. , könnten das Gesuchte enthal-
ten. Gibt es eine Sprache, die einen Artikel aufweist, der aus einem einzigen Vokal besteht und an dem Wort haftet? Da ich zu diesem Zeitpunkt bereits Hebräisch als Sprache des Voynich ver- mutete, lag der Schluss nahe, dass der hebräische Artikel ה ha (mit folgendem Dagesh forte)21 auf den Vokal verkürzt wurde, weil das h nicht gesprochen wurde. Es ist somit zu vermuten, dass die Variante des Hebräischen in einem Land genutzt wurde, wo der Anfangsbuchstabe h nicht ge-
sprochen wird, z.B. im französischen Gebiet. Leider ist diese Überlegung nur zum Teil richtig.

I don't completely agree with this reasoning, since it is based on an assumption. If Hannig wants to find a short article that prefaces words, it's reasonable to also consider the Greek masculine article "o", which Hannig doesn't mention. Even if the VMS turned out to be Hebrew, in the process of researching the text, a person shouldn't ignore patterns that are consistent with other languages.

Quote:Die Zwiebeltürme auf der größten Aus- klappseite verweisen auf eine slawisch-orthodoxe Region.

It doesn't have to be Slavic Orthodox. In European manuscripts, Jerusalem was often drawn with domed roofs, as were Middle-Eastern locations.

There are many possible interpretations of the VMS domes on the rosettes folio. Maybe they are not even domes or towers. They have feet. Maybe they are containers.

He is suggesting that leading "o" represents a noun and leading "4o" represents a verb.

I don't disagree with the possibility that some VMS glyphs might be markers or modifiers. I think that's possible. But the entropy problem becomes worse if they are, so it would be a good idea to address this statistic in the paper.

The paper has a lot more. I don't have time to read the rest now. I'll comment further if I get a chance later.

Hebrew is on my shortlist of possible languages for the VMS, but I envision it being decoded in a way that is different from Hannig's method, so I'm not sure yet if there's any validity to Hannig's translations. I will probably need to read the whole thing twice.

I would like to have seen more than guesswork in his interpretation of gallows and bench-gallows. v101 "sees" seven gallows, EVA only four. Hannig identifies six, and identifies each of his six bench-gallows as a Dagesh. 

I was intrigued by his identification of word-end versions of some letters, which certainly is something others here have discussed, but I'm a little puzzled by his interpretation of EVA-ch as mid- or word-initial EVA-in AND -iin. 

I'm also seeing criticism of the transcriptions and translations on Twitter (my Hebrew is not remotely good enough to undertake that analysis).

It does of course matter how Hannig came to the conclusion that it is Hebrew. What I find much more interesting, however, is whether the tables from page 61 to page 64 are conclusive or not. Unfortunately I lack knowledge of Hebrew to judge this for myself. So it is a question to all who can read it at least to some extent.

(17-06-2020, 11:46 AM)LisaFaginDavis Wrote: You are not allowed to view links. Register or Login to view.I would like to have seen more than guesswork in his interpretation of gallows and bench-gallows. v101 "sees" seven gallows, EVA only four. Hannig identifies six, and identifies each of his six bench-gallows as a Dagesh. 

I fully agree about the amount of guesswork, which is the basis for almost every Voynich solution or explanation, if not all of them.
As far as I can see, many would-be solvers are not all that aware about the number of assumptions they are making (tacit or hidden assumptions).

With respect to the gallows counts, this is quite a bit more complicated   

If one considers characters that appear at least 100 times in the MS, then:
- Eva has 4 plain ones and 3 benched ones
- v101 has 6 plain ones and 3 benched ones

If one considers characters that appear at least 10 times, then:
- Eva has 4 plain ones and 13 benched ones
- v101 has 8 plain ones and 16 benched ones

If one counts all individual characters, then there are literally dozens of all types.

The paper uses the v101 font for character representation.