I've had a chance to look over this proposal, and have come to the opinion that it cannot be correct. I posted my comment at Klaus's site:
http://scienceblogs.de/klausis-krypto-k ... nt-1685988
But I copy it below, also:
I do not speak German, but I could derive certain basic concepts with this work. In my opinion, these clearly show that this cannot be a correct translation of the Voynich. I would categorize them this way:
1) Repeatable? Multiple Hebrew meanings assigned to the same VMs characters; and multiple VMs characters assigned the same Hebrew meaning:
The VMs double "c" can be Hebrew "m" or "n".
The "gallows" can be "k", "t", "p", "ke", "l", and others...
... and so on.
Such variables mean that any translation using them will necessitate and allow a high degree of speculation in determining which ones to use; but is necessary in order to allow the translator to derive any meaning at all from the highly repetitive Voynich. I see this a lot... I mean, the repetitive Voynich demands it, if seen as a language, or as simple substitution for that matter. Any attempt at translation must do one of two things: Allow a high degree of variability to allow a meaningful output, which then ruins any chance of repeatability by others; or the opposite: Accept the repetitive nature of the work, and so accept an output with a high degree of repetition (for example, "duck duck pillow cook duck duck moon pillow sun sun sun"). This current translation falls somewhere in between the two.... giving enough variability to avoid grossly repetitive output, but still shoot for some meaning. Which leads to:
2) Meaning? Even the proposer here has noted that this is problematic. For example, they list some translations:
"1. Certainly, Nymphaea is the twin. Enough juice in the top. Drink
2. careful (?), That's like (something) that delivers mind. Will come with juice
3rd repetition (?). Juice allows speak prophecy ...
4. How to rebel in the presence of prophets
5. Everything in Greek about it is silence without speaking. Bim
6. do not speak (about) juice, said: dig ...
8. ... Spoken in Arabic"
And then they note, "As you can easily see from the lack of words and the blurry meaning, not all difficulties have been solved so far."
So that is an indication the well-meaning translator has seen the problem that everyone will meet: It is so far impossible to balance an output with a "lack of words and the blurring meaning", and this cannot be solved unless one adds FURTHER variability, and ever increasing complexity of "grammar" systems to allow for them. I do admire that they stopped there, but this does fall under one category of attempted translations of the cursed Voynich: The translation proposer realizes the problems, and really has given up and admitted this... but they have not entirely excepted that this is a sign that they are on the wrong track.
In any case, this translation clearly does not fulfill both of the rules I have derived from the various works of the Friedmans, which I paraphrase and simplify into the below. Any translation of any work must fulfill both #1 and #2 at the same time:
1) Repeatable: By someone given a lexicon and set of rules, grammar, and/or system of deciphering, translation and/or decoding, must be able to repeat the results of the proposer.
2) Meaning: The output of above #1 must have meaning in some context.
In my experience, studying dozens of such proposals relating to the Voynich, there are those translations which fall into one of the above categories, but none that come close to falling into both. That would be a solution, and we have not seen one, and this is no exception.
Now the below is not related to this translation proposal, directly, but I point it out because it is one of the many false assumptions, based on poor and incomplete information found on the internet, which causes a great many people to waste precious life-hours, as they begin their investigations on a faulty, error-filled foundation. The author writes (Google Translation) about, and counters, the fake possibility:
"4. A meaningless string of letters in order to simulate a natural language (thus a fake):
"Because of the length of the manuscript and its complexity (illustrations in the style of the 15th century; still slightly schematic plant pictures with roots that signal the medical effect7 etc.), modern counterfeiting (variant c) can be excluded with certainty. This is indicated by the C14 dating to the early 15th century (between 1404 and 1438)."
There are several problems with this: Length and complexity do not "exclude with certainty" a modern fake, as there are many cases of fakes that are long and complex (made for a multitude of purposes). Besides which, any such "complexity" is a speculative concept, as others can analyze the same characters, classifying them in different ways, and see it as a very simplistic, repetitive work. The same with the illustrations: They can speculatively be deemed accurate or not; complex, or not; genuine or fanciful; of high quality, or not, and so on. So to base any judgement on the era and/or authenticity of the work can only be based on a highly variable and speculative assessment of the text and illustrations.
Also, the C14 dating does not date the Voynich, it only dates the calfskin it is constructed from. There are many cases in which parchment, paper, and other writing materials have been used up to hundreds of years after they were constructed. As a related note, Wilfrid himself was one of the best-placed persons to have access to such materials, as he purchased the Libreria Franceshini in 1908, which contained up to a half a million items of every type, including scrap.
But in short, I would say that we know little with "certainty", most of all those things which go beyond the few "certains" we do know, and into the realm of opinion. But again, the problem (as I see it) is that all new Voynich researchers are saddled with a set of such claimed "expertly determined" absolutes, which they will not have the time, energy, nor resources to take, to set their own, more accurate foundations, on which to start their work. I won't use the terminology of the famous phrase about computer programming, as I in no way want to impugn the fine attempt of this earnest proposer, but rather I will paraphrase it, "If you use bad information to input; you will only get faulty results".