+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Analysis of the text (https://www.voynich.ninja/forum-41.html)
+--- Thread: Voynich text in other characters (/thread-2752.html)
Voynich text in other characters - Mark Knowles - 18-04-2019
One suggestion I would make if it hasn't already been made is to try and generate Voynichese like text with a different invented alphabet, i.e. set of symbols, as I think using the Voynich alphabet can be confusing as on the one hand you are trying to produce text is structured like the Voynich text, but is not exactly identical to Voynich text in its structure. This is clearly not an attempt to directly decipher the Voynich, but rather to simulate it.
RE: Voynich text generation - -JKP- - 18-04-2019
Quote:Mark Knowles: One suggestion I would make if it hasn't already been made is to try and generate Voynichese like text with a different invented alphabet, i.e. set of symbols, as I think using the Voynich alphabet can be confusing as on the one hand you are trying to produce text is structured like the Voynich text, but is not exactly identical to Voynich text in its structure. This is clearly not an attempt to directly decipher the Voynich, but rather to simulate it.
If we are trying to simulate Voynich text, it makes sense to use Voynich glyphs.
It seems to me this is less rather than more confusing because it makes it easier to spot the parts that are not exactly identical to Voynich text (the whole point is to analyze these small differences in order to better understand the VMS structure).
RE: Voynich text generation - Mark Knowles - 18-04-2019
My perspective is that unless someone has almost solved the Voynich cipher it is better to try to produce a Voynich like cipher which exhibits a similar structure without being bound by the Voynich symbols, because of the confusion that can result.
From discussions that I have had I get the impression that nobody is close to solving the cipher, so using Voynich symbols is unhelpful.
RE: Voynich text generation - Koen G - 18-04-2019
Like JKP I don't understand how this would help, Mark. If you want to see how to approach Voynichese patterns, doing so with other symbols would add a very confusing step.
RE: Voynich text generation - Mark Knowles - 18-04-2019
Koen: I suppose with one's own cipher and set of symbols one could still look at questions of entropy,, repeated words and all other general properties of Voynich text without having to match how each symbol appears exactly in Voynichese. In a sense not being bound by Voynich symbols one can explore the kind of cipher that we have without the extra complication of having to fit the precise behaviour of each symbol, matching in to the way it in the Voynich. To me it seems like a good intermediate step, but of course not the end goal. So this means:
Step 1) Construct a cipher in an original alphabet of a similar number of characters to Voynichese which has the same general statistical or other properties as Voynichese.
If Step 1 successful Then
Step 2) Using the Voynich alphabet construct a cipher of a similar kind to step 1, which precisely matches what we see in Voynichese.
I think by breaking these up into 2 separate steps or tasks it makes the end goal easier to arrive at.
In step 1 we establish the general principles and in step 2 will drill down to the specifics. By using an original alphabet in Step 1 it is much easier to detach it from Step 2.
This is the approach I am adopting as I think it is more convenient and easier.
RE: Voynich text generation - Mark Knowles - 18-04-2019
Koen: I can produce my own cipher say "Knowleses" with my own Knowleses glyphs and that Knowleses mimics the kind of properties which we see in Voynichese then I can say my Knowleses looks like a close cousin to Voynichese. If I can achieve this stage of producing a Voynich-like cipher than I will be ready to tackle Voynichese itself. I believe in breaking up problems into smaller steps as it makes them more managable and so easier to solve.
RE: Voynich text generation - -JKP- - 18-04-2019
(18-04-2019, 03:03 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.Koen: I suppose with one's own cipher and set of symbols one could still look at questions of entropy,, repeated words and all other general properties of Voynich text without having to match how each symbol appears exactly in Voynichese. In a sense not being bound by Voynich symbols one can explore the kind of cipher that we have without the extra complication of having to fit the precise behaviour of each symbol, matching in to the way it in the Voynich. To me it seems like a good intermediate step, but of course not the end goal. So this means:
...
Certainly there are situations where one doesn't want to be bound by Voynich symbols, but this thread was a very specific exercise (a good idea, in my opinion) and as Koen pointed out, and I agree, it would add an additional layer of complication, and that is simply not necessary to achieve the goals posed by the exercise.
Feel free to make the suggestion for alternate characters on other threads. I honestly don't think it's appropriate here, but I'll defer to Rene, since he posed the exercise.
RE: Voynich text generation - Mark Knowles - 18-04-2019
I posted it on this thread as I thought it relevant to the topic at hand, but if someone thinks it belongs on another thread that is fine by me. I stand by my point, but I am not concern where that point appears.
RE: Voynich text generation - ReneZ - 18-04-2019
The problem with presenting text in the Voynich script is that it automatically makes the text look 'realistic' even if perhaps it isn't. This was evident when Gordon Rugg first presented his theory about the Cardan Grille method. The generated text really looked like 'Voynichese' to the more general public, but not to those who were very familiar with it.
The way to avoid this is to use one of the common transcription alphabets. I don't see any advantage in using other symbols.
RE: Voynich text generation - -JKP- - 18-04-2019
(18-04-2019, 04:34 PM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.The problem with presenting text in the Voynich script is that it automatically makes the text look 'realistic' even if perhaps it isn't. This was evident when Gordon Rugg first presented his theory about the Cardan Grille method. The generated text really looked like 'Voynichese' to the more general public, but not to those who were very familiar with it.
...
I completely agree that Rugg's text doesn't look at all like Voynichese. One can see it at a glance. In fact, it's farther off than one would expect for a person trying to prove a point about how Voynich text was originally generated.