The Voynich Ninja

(15-02-2016, 09:36 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(15-02-2016, 04:30 AM)-Job- Wrote: You are not allowed to view links. Register or Login to view.It can be difficult to separate a cipher from a constructed language.

Maybe a better question would be, why is the Voynich Manuscript not a natural language?

It's really not better to phrase it that way. I want positive reasons for the Voynich manuscript being a cipher. If folks' beliefs truly boil down to, "it's a cipher because I don't think it's a language", then I think the whole side of cipher research on the VM has a serious problem.

At the same time, asserting that "it's a language because nobody can prove to me beyond all reasonable doubt that it's a cipher" would be doubly wrong. :-)

But anyway, we do have plenty of examples of things in the Voynich that directly point to its being a cipher. Here are a few:

(1)  Because "aiiv" and "aiir" groups / families are appropriated from a highly stylized and well-known existing scribal practice (medieval page references - quire / folio / side), there is a high expectation - if Voynichese is a language - that these should function in the same way that medieval page numbers do. Yet they do not - not even slightly. This implies that there is a disjunct between the appearance of the text and the content of the text: which is the very definition of cryptography.

(2) In the 15th century, the "4o" shape appears in a number of (pre-Arabic number) Northern Italian cipher alphabets: yet I am not aware of any 15th century texts in which it appears. If you accept that the VMs is a product of the 15th century, then you surely have to accept that the place you are most likely to find this shape is in cipher documents.

(3) The "or" / "ar" / "ol" / "al" patterns occur overwhelmingly as pairs within the Voynichese text: yet there are places where the spaces appear to have been shifted around. Look at f15v: there, the first line’s “ororor” has been turned into “oror or“, while the second line’s “orororor” has been turned into “or or oro r“. I don't believe this is a copying error: I believe the space has been shuffled around to conceal the "or or or or" structure, to make it less obvious to the eye. And given that this is a cipher trick, this suggests that the whole thing is a cipher.

(4) The presence of a systematic set of artificial characters (gallows) that is not attested anywhere else points to the presence not only of an invented alphabet, but also of a systematic invented alphabet, which is normally a hallmark of cipher alphabets.

(5) The complete absence of structural matches between Voynichese and conventional languages - despite extensive searching by a large number of people - points to a fundamental problem with linguistic interpretations. This is, I believe, primarily what pushed Friedman in the direction of artificial languages, even though these were incompatible with the dating evidence. If you want to believe that Voynichese is a language, then go ahead: but the evidence of a century of searching points to zero evidence for a real-world language. And if you do go down that route, you will inevitably end up making a complete and utter Bax of yourself... so be careful.

(6) Even though Voynichese looks like a language, this does not mean it is not a cipher: because the idea of removing spaces from ciphers was not proposed until the 16th century, ciphertexts from this period almost all looked like languages too. (Though we have evidence from Alberti that transposition ciphers were discussed in the 15th century, we have no actual example of their being used for real).

I'll agree that a complex cypher is improbable around the time the VMS was made. The ones I've seen from that time period are pretty straightforward.

But what Mozart did was improbable too—composing symphonies at the age of four. It's possible he was terrible at art, terrible at business, terrible at business, terrible at many things, but he had a gift for music. Maybe the VMS author had a gift for certain kinds of symbolic thinking.

(15-02-2016, 12:24 PM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.I'll agree that a complex cypher is improbable around the time the VMS was made. The ones I've seen from that time period are pretty straightforward.

The notion that Voynichese necessarily uses a complex cipher is just a presumption - until we decrypt it, we won't know if it is complex or just cunning (i.e. several steps sideways from what we are expecting). :-)

(15-02-2016, 09:36 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(15-02-2016, 04:30 AM)-Job- Wrote: You are not allowed to view links. Register or Login to view.It can be difficult to separate a cipher from a constructed language.

Maybe a better question would be, why is the Voynich Manuscript not a natural language?

It's really not better to phrase it that way. I want positive reasons for the Voynich manuscript being a cipher. If folks' beliefs truly boil down to, "it's a cipher because I don't think it's a language", then I think the whole side of cipher research on the VM has a serious problem.

I agree that it's important to list supporting features (which i did), my point is that ciphers and constructed languages overlap.

For example, the VM's internal word structure could be taken as evidence of a cipher or a constructed language, so it won't make it into the list of reasons why it's a cipher.

The line is clearer between a natural or artificial text.

Out of curiosity, what positive arguments do you have for believing the text is an unencoded natural language?

"...Even though Voynichese looks like a language, this does not mean it is not a cipher..."

If the "o" characters were nulls and the "a" characters represent consonants and the spaces between word-tokens were contrived (I don't want to use the word "arbitrary" because they're not) then it wouldn't look very much like a language. It's the positioning of the Latin "a" and "o" that give it a perhaps deceptive familiarity.

I can think of some simple steps that can produce a really tough cipher.

For example, suppose the author encodes the following text:

The Rain in Spain Falls Mainly on the Plain

As follows:

De rhein een cepain faals mayneli oen the pelain.

Then writes it down using a different alphabet. The result would be essentially as practical as a simple monoalphabetic cipher but much tougher to break because it's less susceptible to frequency analysis (among others).

As Nick said, we don't know if the cipher is complex or just cunning.

(15-02-2016, 12:24 PM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.I'll agree that a complex cypher is improbable around the time the VMS was made. The ones I've seen from that time period are pretty straightforward.

But what Mozart did was improbable too—composing symphonies at the age of four. It's possible he was terrible at art, terrible at business, terrible at business, terrible at many things, but he had a gift for music. Maybe the VMS author had a gift for certain kinds of symbolic thinking.

That simile doesn't work. Mozart was just exploring something that already existed, he didn't "create" new musical notes.
I can accept that the VM is a cunning cipher. But I doubt very much that the Scribe managed to leapfrog several hundred years of linguistic, semiotic and mathematical development to create a new type of unbreakable cipher. He just didn't have the tools at his disposal to make that sort of breakthrough.

(20-02-2016, 05:44 AM)-Job- Wrote: You are not allowed to view links. Register or Login to view.I can think of some simple steps that can produce a really tough cipher.

For example, suppose the author encodes the following text:

The Rain in Spain Falls Mainly on the Plain

As follows:

De rhein een cepain faals mayneli oen the pelain.

Then writes it down using a different alphabet. The result would be essentially as practical as a simple monoalphabetic cipher but much tougher to break because it's less susceptible to frequency analysis (among others).
...

It might be much harder to decipher by frequency analysis, but it would not be very hard to decipher for the human brain. I recently deciphered a simple substitution code in an old manuscript by just looking at and I'm in no way a cryptographer. Just changing the shape of the letters doesn't change the function of the letters.

Changing "The" to "De" isn't a hindrance, since the Anglo-Saxon and Norse "d" character with the bar or the Greek theta immediately come to mind for polyglots and its position in the word makes it a high candidate as a definite article, regardless of how it's spelled. The short "word" in front of the fairly short last word makes it unlikely that it's certain languages that put the verb at the end. Lots of clues.

Your example retains the grammatical structure and the basic vowel-consonant balance at the beginning and end of each word, even if it has extra vowels inserted midword. Italian, for example, has extra vowels too, compared to some languages, so that detail is only a minor hindrance to decryption.

It might be difficult for someone who knows only one language to decipher but a polyglot who is used to switching languages in his or her head could probably make it out without any computer assistance. The biggest problem is not the extra letters, but the small size of the sample. The less text there is, the less opportunity there is to see patterns and how individual parts of the words behave.


If the above words were swapped around so the sentence structure is changed, then it becomes much more difficult. It would obscure the grammatical pattern. The problem for the encypherer is to do it in such a way that it can be decrypted/reconstructed. Cryptography is like bad handwriting. If you write it so you can't read it back, it's useless.

(20-02-2016, 05:44 AM)-Job- Wrote: You are not allowed to view links. Register or Login to view.De rhein een cepain faals mayneli oen the pelain.

In fact, this type of alternative and inconsistent spelling (though maybe not quite as adventurous) should probably be expected.
I wouldn't call it part of a cipher....

When I occasionally try to read German texts from 15th-16th Century books and MSs, they come across as something like the above, compared to modern German.

(20-02-2016, 10:24 AM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.Changing "The" to "De" isn't a hindrance, since the Anglo-Saxon and Norse "d" character with the bar or the Greek theta immediately come to mind for polyglots and its position in the word makes it a high candidate as a definite article, regardless of how it's spelled. The short "word" in front of the fairly short last word makes it unlikely that it's certain languages that put the verb at the end. Lots of clues.

The substitution of "the" with "de" would not remain consistent throughout the text, that's a key point. In my example i used both "the" and "de" to illustrate this, and would have certainly used additional variants in a longer text, with other substitutions as well as null characters (e.g. 'h').
You could find a way in via grammatical analysis (as you mentioned) and lots of trial and error, but that's assuming you're not wasting time with other alternatives.