The Voynich Ninja

Full Version: How perfect is Voynichese?
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5
[quote="Mauro" pid='87368' dateline='1783632694']A You are not allowed to view links. Register or Login to view. is not a codebook cipher. The resulting encrypted text is a random string and gives zero informations to an attacker.[quote]

Ah, so you mean a Vigenère-type cipher.  I though @JoJo was referring to a codebook-style cypher, namely a substitution cipher operating on words instead of characters.  Even if the codebook is used once and then lost or discarded, if the text is long enough...

I think that a Vigenère-type cipher is rather unlikely because it would "homogenize" the word frequencies -- instead of a "most common word" there would be multiple equally common "most common words", etc.  So it would have to be someting even more complicated than a simple Vigenère.  Seems rather unlikely for the times...

All the best, --stolfi
[quote="Jorge_Stolfi" pid='87381' dateline='1783649445']
[quote="Mauro" pid='87368' dateline='1783632694']A You are not allowed to view links. Register or Login to view. is not a codebook cipher. The resulting encrypted text is a random string and gives zero informations to an attacker.[quote]

Ah, so you mean a Vigenère-type cipher.  I though @JoJo was referring to a codebook-style cypher, namely a substitution cipher operating on words instead of characters.  Even if the codebook is used once and then lost or discarded, if the text is long enough...

I think that a Vigenère-type cipher is rather unlikely because it would "homogenize" the word frequencies -- instead of a "most common word" there would be multiple equally common "most common words", etc.  So it would have to be someting even more complicated than a simple Vigenère.  Seems rather unlikely for the times...

All the best, --stolfi
[/quote]
Not necessarily complicated (and not comparable to the Visener), but simply redundant. In the case of the Visener, we can apply the key and quickly get the text, but according to my version, the VMS cipher does not have mathematical rigor, and it does not have any specific algorithm, encryption could be done manually.
(10-07-2026, 09:42 AM)ololololo Wrote: You are not allowed to view links. Register or Login to view.Not necessarily complicated (and not comparable to the Visener), but simply redundant. In the case of the Visener, we can apply the key and quickly get the text, but according to my version, the VMS cipher does not have mathematical rigor, and it does not have any specific algorithm, encryption could be done manually.

Could you explain simply in 2-3 sentences how the VMS cipher works in your hypothesis? Maybe even if you could show it with an example, suppose we want to encode "WHAT THE MANUSCRIPT SAYS ABOUT THE PLANTS AND STARS IS A RIDDLE WRAPPED IN A MYSTERY INSIDE AN ENIGMA", how would we do this?
[quote="Jorge_Stolfi" pid='87381' dateline='1783649445']
[quote="Mauro" pid='87368' dateline='1783632694']A You are not allowed to view links. Register or Login to view. is not a codebook cipher. The resulting encrypted text is a random string and gives zero informations to an attacker.[quote]

Ah, so you mean a Vigenère-type cipher.  I though @JoJo was referring to a codebook-style cypher, namely a substitution cipher operating on words instead of characters.  Even if the codebook is used once and then lost or discarded, if the text is long enough...

I think that a Vigenère-type cipher is rather unlikely because it would "homogenize" the word frequencies -- instead of a "most common word" there would be multiple equally common "most common words", etc.  So it would have to be someting even more complicated than a simple Vigenère.  Seems rather unlikely for the times...

All the best, --stolfi
[/quote]

Well, one-time pads are Vigenere ciphers where the key is as long as the plaintext and truly random and  never reused. They are on a completely different planet from Vigenere's: one-time pads are unbreakable, even with infinite computer power.

Of course I'm not suggesting one-time pads have anything to do with the VMS! It was only to correct a sentence in a previous post from the OP: "anagrams + substitutions is about as safe as a one-time pad". Not even close.
(10-07-2026, 10:01 AM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.
(10-07-2026, 09:42 AM)ololololo Wrote: You are not allowed to view links. Register or Login to view.Not necessarily complicated (and not comparable to the Visener), but simply redundant. In the case of the Visener, we can apply the key and quickly get the text, but according to my version, the VMS cipher does not have mathematical rigor, and it does not have any specific algorithm, encryption could be done manually.

Could you explain simply in 2-3 sentences how the VMS cipher works in your hypothesis? Maybe even if you could show it with an example, suppose we want to encode "WHAT THE MANUSCRIPT SAYS ABOUT THE PLANTS AND STARS IS A RIDDLE WRAPPED IN A MYSTERY INSIDE AN ENIGMA", how would we do this?
In fact, I can only demonstrate how it roughly works.
First, let's create a cipher alphabet: A = I, B = II... Z = XXV (U=V=XXI).
We encrypt and align the letters in order:
XXII, XX, VIII, I / XX, VIII, V / XXI, XX, XIX, XVIII, XVI, XIV, XIII, IX, III, I / XXIV, XIX, XIX, I / XXI, XX, XV, II, I / XX, VIII, V / XX, XIX, XVI, XIV, XII, I / XIV, IV, I / XX, XIX, XIX, XVIII, I / XIX, IX / I / XVIII, XII, IX, V, IV, IV / XXII, XVIII, XVI, XVI, V, IV, I / XIV, IX / I / XXIV, XXIV, XX, XIX, XVIII, XIII, V / XIX, XIV, IX, IX, V, IV / XIV, I / XIV, XIII, IX, VII, V, I

We introduce a small alphabet: I = i, V = a, X = e, XX = ch (if separately, then sh), XV = o, IV = ok/ot, IX = ol/or, VI = da; I = ol, II = or (XXII = ch-or)
chor sh daiin i / sh daiin a / chol sh eor edaiin oi eol eiin ol iin i / chok eol eol i / chi sh ea ii i / sh daiin a / sh eor oi eok ein i / eot ot i / sh eor eor oiin i / eor or / i / oiin ein or a ia a / chin oiin oi oi a ia i / eot or / i / chok chot sh eol oiin eiin a / eol eok ol or a ia / eok i / eok eiin ol ain (I encrypted it manually, there may be errors. The last i = n).

To decipher this text, you first need to substitute the first alphabet, then the alphabet with numbers and solve the anagrams. You can assign this task to a computer. In my version, this encryption is called primary.
But let's try to embellish this text a little, for example, combine beautiful words a la sheor and use VMS symbols that are not in the alphabet:
chor sh daiin im / shedaiin am / chol sheor edaiin oi eol eiin iin im / chokeo l eol im / chi sh eaiin/ sh daiin am / sheor oi eo kein im / eot ot im/ sheor eor oiin im / eor or / otchy im / oiin ein ora ia am / chin oiin oioia iaim / e otor / kaiin im / chok chot sheol oiin eiin am / eol eok olor a iam / eok i / eok eiin olain /

I chose m for the end of the line, and in the parts where there was only i, I added one meaningless word each. In my theory, this is called secondary encryption.
This is actually a very, very simple example. It is unlikely that you will have difficulty deciphering the second text. But I can complicate it, for example, by adding more meaningless words or meaningless letters (in order to eliminate type ia deformities). In the process, I can create so many internal rules that you, knowing the alphabet but not knowing the rules I use, will never be able to decipher this text.

(10-07-2026, 10:47 AM)Mauro Wrote: You are not allowed to view links. Register or Login to view."anagrams + substitutions is about as safe as a one-time pad"
Don't take everything so literally Big Grin
(10-07-2026, 11:05 AM)ololololo Wrote: You are not allowed to view links. Register or Login to view.In fact, I can only demonstrate how it roughly works.
...
This is actually a very, very simple example. It is unlikely that you will have difficulty deciphering the second text. But I can complicate it, for example, by adding more meaningless words or meaningless letters (in order to eliminate type ia deformities). In the process, I can create so many internal rules that you, knowing the alphabet but not knowing the rules I use, will never be able to decipher this text.

Thank you, now I understand how the scheme works. Deciphering it can be problematic, but I'm not sure it's as unbreakable as you suggest, especially for a very long text.
(10-07-2026, 12:04 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.
(10-07-2026, 11:05 AM)ololololo Wrote: You are not allowed to view links. Register or Login to view.In fact, I can only demonstrate how it roughly works.
...
This is actually a very, very simple example. It is unlikely that you will have difficulty deciphering the second text. But I can complicate it, for example, by adding more meaningless words or meaningless letters (in order to eliminate type ia deformities). In the process, I can create so many internal rules that you, knowing the alphabet but not knowing the rules I use, will never be able to decipher this text.

Thank you, now I understand how the scheme works. Deciphering it can be problematic, but I'm not sure it's as unbreakable as you suggest, especially for a very long text.
In this case, the difficulty may be in the lack of information and the reconstruction of the alphabet, but apparently the latter is still possible to do.
I've looked through past proposals that involved Roman numerals, the following one is very close, but it actually makes it even more destructive by summing the values of letters in each word and writing the result as a single Roman numeral (and then mapping to Voynichese glyphs). However, to make it a bit easier to tell apart different words, the letter values are not 1-23, but larger. The result actually looks quite similar to Voynichese:
You are not allowed to view links. Register or Login to view.

(14-09-2024, 12:04 AM)RobGea Wrote: You are not allowed to view links. Register or Login to view.Some letter values:
t = 116
h = 104
e = 101
c = 99
a = 97

PLAINTEXT = "the cat sat on the mat"

the :: 321 :: [100, 100, 100, 10, 10, 1] 321 ['C', 'C', 'C', 'X', 'X', 'I']
cat :: 312 :: [100, 100, 100, 10, 1, 1] 312 ['C', 'C', 'C', 'X', 'I', 'I']
... etc ...

INTERMEDIATE = "CCCXXI CCCXII CCCXXVIII CCXXI CCCXXI CCCXXII" 

CIPHERTEXT = "chkem qoshol qokeriin qoem qokem qokeol"
(10-07-2026, 05:04 PM)oshfdk Wrote: You are not allowed to view links. Register or Login to view.I've looked through past proposals that involved Roman numerals, the following one is very close, but it actually makes it even more destructive by summing the values of letters in each word and writing the result as a single Roman numeral (and then mapping to Voynichese glyphs). However, to make it a bit easier to tell apart different words, the letter values are not 1-23, but larger. The result actually looks quite similar to Voynichese:
You are not allowed to view links. Register or Login to view.

(14-09-2024, 12:04 AM)RobGea Wrote: You are not allowed to view links. Register or Login to view.Some letter values:
t = 116
h = 104
e = 101
c = 99
a = 97

PLAINTEXT = "the cat sat on the mat"

the :: 321 :: [100, 100, 100, 10, 10, 1] 321 ['C', 'C', 'C', 'X', 'X', 'I']
cat :: 312 :: [100, 100, 100, 10, 1, 1] 312 ['C', 'C', 'C', 'X', 'I', 'I']
... etc ...

INTERMEDIATE = "CCCXXI CCCXII CCCXXVIII CCXXI CCCXXI CCCXXII" 

CIPHERTEXT = "chkem qoshol qokeriin qoem qokem qokeol"
Quite interesting. Perhaps something like this is used in Currier B, given the abundance of qo-. But at the moment, I can't say anything here...
It seems that chkem qoshol qokeriin qoem qokem qokeol falls short of perfection due to e-i combinations.
Pages: 1 2 3 4 5