The Voynich Ninja

(22-01-2024, 05:33 PM)dfs346 Wrote: You are not allowed to view links. Register or Login to view.It seems to me that the proposed mapping between Voynich glyphs and Persian letters should address the frequencies of the respective symbol sets.

If we use the v101 transliteration, herbal section, as the corpus of reference, as illustrated in the tables below, there is in most cases a large difference between the frequency of the Voynich glyph (or glyphs) and that of the Persian letter (or letters) proposed to correspond. 

The Persian letter frequencies are from a corpus of modern Persian, and the frequency analysis would be better if based on Middle Persian (for example the Ruba'iyat of Omar Khayyam); but I'm inclined to think that the frequencies in Middle Persian would not be much different.

I just noticed here some vowels from Table 1 are confused with other vowels/letters, for A it should be ـَ and for V it should be ـُ
ـُ is basically an O, while ء is pronounced like ae similar to ع

Here's my attempt at You are not allowed to view links. Register or Login to view. line 2, using the proposed mapping. I did not see any repetitions on this line, so did not apply rule 1. (Unless v101-ho8 {rids} or v101-89 {sv} are considered repetitions; it wasn't clear to me whether removal of repetitions applies to any repeated string, or only to isolated strings or "words".)

[attachment=8105]

(23-01-2024, 08:52 AM)dfs346 Wrote: You are not allowed to view links. Register or Login to view.Many thanks for this clarification. Indeed, we don't need to use the v101 transliteration; we can map directly from the glyphs to Persian letters, as per the proposed mapping. That would be a manual task. The advantage of using v101 as an intermediate step is that the v101 glyphs are available in Unicode, as are the Persian letters. Therefore, using the find-and-replace function in Microsoft Word or Excel, it's a half-hour's work to apply the proposed mapping to a line, a paragraph, a page or even the whole Voynich manuscript.


I attempted to apply the proposed mapping, followed by rules 1 and 2, to page You are not allowed to view links. Register or Login to view. line 1. I used Unicode symbols for some of the phonetic strings, to allow the find-and-replace function to work, as follows:

• all upper case mapped to lower case
  
• A' mapped to á
  
• T' mapped to ť
  
• T" mapped to ŧ
  

This yielded the following result:


I noted your interpretation (Step 3) and copied it into Google Translate, with the following result:


I would like to invite the comments of this forum on the extent to which Step 3 is arbitrary.

Google-translate is translating حوادث correctly, but other words require dictionary search. There is no verb "to be" in these two noun phrases. I like to explain that some crazy things are happening:
Two phrases are almost in one meaning:
(You are not allowed to view links. Register or Login to view.) translates to increase and growth, (You are not allowed to view links. Register or Login to view.) means growth, mostly used for growth of plants as (نبات) defines for plant itself (dictionary results from abadis.ir are linked, you can enable google translate in browser so it will translate the definitions to some degree)
(You are not allowed to view links. Register or Login to view.) means incidents in negative form, (You are not allowed to view links. Register or Login to view.) means incident in almost the same meaning.

So we have "growth of incident growth of incidents "

(23-01-2024, 10:41 AM)dfs346 Wrote: You are not allowed to view links. Register or Login to view.Here's my attempt at You are not allowed to view links. Register or Login to view. line 2, using the proposed mapping. I did not see any repetitions on this line, so did not apply rule 1. (Unless v101-ho8 {rids} or v101-89 {sv} are considered repetitions; it wasn't clear to me whether removal of repetitions applies to any repeated string, or only to isolated strings or "words".)

In the method, throughout a complete paragraph, all words undergo changes. Of course, you can check it in a small sequence, but the effect of deleting the previous and next iterations will be lost. For this sequence you have selected (without considering the previous sequence):

vpttv a'ar dr da'v sa'd t'ridsv dridsar a'dsv

vptv a'a  a'v sa'd t'ridsv dridsar a'dsv

vptv a'a  a'v sa'd t'ridsv dridsar a'dsv

vptv a v sa'd t'ridsv dridsar a'dsv

There are many combinations like the part which turned to و درید سار, it may be the following words: و = and, درید = to cut, to tear (verb), and سَر =head (noun).

ُsome of the words in your attempt: You are not allowed to view links. Register or Login to view. = lentil, You are not allowed to view links. Register or Login to view.= auspicious (سعد also has a plant associated defined as سعد خوراکی).

Hello Pardis
This is quite complicated what you present us here.

So my simple question would be - are you suggesting that Voynich script is a basic substitution cipher or something more?

By basic substitution I mean that any letter is always replaced by the same sign, like in Ceasar's cipher.
Or is it something more when it comes to Voynich?

You mention some patterns of repetition but I don't quite get it. Do they change the meaning of symbols?

(24-01-2024, 01:57 PM)Rafal Wrote: You are not allowed to view links. Register or Login to view.Hello Pardis
This is quite complicated what you present us here.

So my simple question would be - are you suggesting that Voynich script is a basic substitution cipher or something more?

By basic substitution I mean that any letter is always replaced by the same sign, like in Ceasar's cipher.
Or is it something more when it comes to Voynich?

You mention some patterns of repetition but I don't quite get it. Do they change the meaning of symbols?

Hi Rafal
No, I suggest the creator of this manuscript, in addition to the common methods, has added disturbing parts to the cuttings of the text with specific rules. The spaces and words in the lines are not real, and these pseudo-lines are not meant to be translated into something comprehensible. Assuming we have a correct alphabet set, if we read the text, we will not understand anything from it, or we will not understand it completely. So he kept his knowledge from some people with this code. A layer-by-layer code since by removing each type of these repetitions, or disturbing elements, new phrases and words will appear.

So can you explain with an example, step-by-step, how the author converted a sentence of correct Persian into Voynichese, and how the reader then converts Voynichese back into the correct Persian?

What is unclear to me in the examples in your colored cells PDF is the order for applying the omission rules. In some cases like the first paragraph of You are not allowed to view links. Register or Login to view. you have all three colors at the same time: type one (magenta), type two (blue), type three (green).

Quote:Within a word, two elements B and B* are same surrounding a specific region:

The type 3 rule is unclear too: what does "a specific region" mean?

Sometimes, the result depends on the order you apply the rules, and whether removals are done in a second pass, after identification of the repetition patterns or if they are done in a single pass (search, remove, repeat): then the direction of search, left to right or right to left matters.

Identification then removal:
ay by cy -> a b c

Single pass, left to right:
ay by cy -> a b cy

Looking at your colored cells PDF it seems to be the former.

Quote:The order of deletions was according to the number of the rules in such a way that first the repetitions of type 1 were reduced to one, then frequent repetitions of type 2 were removed. Finally, type 3 repetitions were removed from the remaining text. This order of removing repetitions continued until no form of repetition remained.

Type 1, then 2, then 3, then loop. Got it I think.

Ok, so if I understand correctly the spaces are not real spaces but a way to hide and obfuscate the real meaning.

Well, it's possible and if I recall correclty it was considered by some people before.

And it was used in practice, after all Wikipedia says in its basic article about ciphers:

Usually the ciphertext is written out in blocks of fixed length, omitting punctuation and spaces; this is done to disguise word boundaries from the You are not allowed to view links. Register or Login to view. and to help avoid transmission errors. These blocks are called "groups", and sometimes a "group count" (i.e. the number of groups) is given as an additional check. Five-letter groups are often used, dating from when messages used to be transmitted by You are not allowed to view links. Register or Login to view.:

SIAAZ QLKBA VAZOA RFPBL UAOAR

But there must be some fixed rule how to remove this spaces.
If you start to join and break vords (Voynich words) free at will then your solution becomes very ambiguous and two different people reading the same text may get a very different solution.

(26-01-2024, 11:44 AM)nablator Wrote: You are not allowed to view links. Register or Login to view.What is unclear to me in the examples in your colored cells PDF is the order for applying the omission rules. In some cases like the first paragraph of You are not allowed to view links. Register or Login to view. you have all three colors at the same time: type one (magenta), type two (blue), type three (green).

Quote:Within a word, two elements B and B* are same surrounding a specific region:

The type 3 rule is unclear too: what does "a specific region" mean?

Sometimes, the result depends on the order you apply the rules, and whether removals are done in a second pass, after identification of the repetition patterns or if they are done in a single pass (search, remove, repeat): then the direction of search, left to right or right to left matters.

Identification then removal:
ay by cy -> a b c

Single pass, left to right:
ay by cy -> a b cy

Looking at your colored cells PDF it seems to be the former.

Quote:The order of deletions was according to the number of the rules in such a way that first the repetitions of type 1 were reduced to one, then frequent repetitions of type 2 were removed. Finally, type 3 repetitions were removed from the remaining text. This order of removing repetitions continued until no form of repetition remained.

Type 1, then 2, then 3, then loop. Got it I think.

Loop ( loop(Type 1), then loop (Type2), then loop(Type3) )

ay by cy -> a b cy certainly gives different results from ay by cy -> a b c. However, the form I see is something like: aty bty cy so we can see there are two layers of repetitions, and ultimate goal is to remove these disturbing elements. Here these elements go with rule 2, aty bty cy -> at bt c then at bt c -> a b c. It requires a loop for rule 2 and should start again from beginning of the paragraph each time after removing. So that, if y is removed the t won't be left behind.

By specific region I mean a letter or a sequence of letters inside a word, if we imagine a symmetry line, two repeated elements are placed in two sides of the inner segment (sometimes even mirrored).