The Voynich Ninja

I just watched the short video on the VMS created by TED-ED and scripted by Stephen Bax.

The graphics are charming. I enjoyed them, and they'll probably reach a broader audience and generate increased interest in the VMS.


There is, however, a statement in the video that I find questionable, and since this section is for analysis of the text, I thought it might be interesting to discuss it.

Here is the statement:

"Cryptologists say the writing has all the characteristics of a real language."


Is that really true? Is that the general consensus among cryptologists? Is there a general consensus among cryptologists?


I'm not a cryptologist, but my personal feeling is that the text diverges quite significantly from natural language, particularly if the spaces are taken literally. I have never seen any natural language with such a high level of repetition combined with such a high degree of positional rigidity, not only in the word-tokens themselves, but also in the construction of individual lines.

The only way I can see this being natural language (and I still haven't discounted the possibility) is IF the spaces are contrived AND there are nulls AND some of the letter forms that appear similar are possibly intended to be differentiated by details (e.g., the length or direction of tails) OR if it is another kind of system entirely (musical notes, numbers, semaphore, etc., which is then converted back into natural language).


Your thoughts?

Mod edit: The link to the video is You are not allowed to view links. Register or Login to view..

Wishful thinking and marketing.

Logically, if we can say with assurance that it has "all the characteristics of a real language" then we have matched it with a real language and can start the decryption process.

I assume they're extrapolating from the Zipf law and all the rest of the assumptions surrounding it, but that's a mathematical assumption, not a crypto-graphical one.

Reference is clearly to the Montemurro and Zanette paper.

It's also clear that Stephen was under certain constraints, he wasn't allowed to refer to specific research etc. On the other hand it's true that there seems to be some self-promotion involved, with the Tauron and such. But that's good for the general public.

The problem with the "natural language" question is that there are so many possibilities. Maybe we can statistically prove that it's not full sentences in any normal language. But could it be lists of nouns/names/adjectives? Directions, measurements, instructions...? In other words, might it be natural language in an unexpected text type?

There's another possibility that's been wandering through my mind lately, but it's hard to explain and it may be a dumb idea. I'll make a new thread 

(26-05-2017, 07:30 PM)Koen Gh. Wrote: You are not allowed to view links. Register or Login to view.It's also clear that Stephen was under certain constraints, he wasn't allowed to refer to specific research etc. On the other hand it's true that there seems to be some self-promotion involved, with the Tauron and such. But that's good for the general public.

The problem with the "natural language" question is that there are so many possibilities. Maybe we can statistically prove that it's not full sentences in any normal language. But could it be lists of nouns/names/adjectives? Directions, measurements, instructions...? In other words, might it be natural language in an unexpected text type?

There's another possibility that's been wandering through my mind lately, but it's hard to explain and it may be a dumb idea. I'll make a new thread 

I've spent many hours trying to reconcile the level of repetition and rigidity and I can't even get poetry, abbreviations, and lists to fit a format as unusual as the VMS, at least in its raw form. It would have to be lists of the same things and only of items with unusually similar names.

I think it's a big mistake to try to pull out words from labels or from individual parts of lines and trust the interpretation without broader corroboration, as these approaches have universally failed to generalize to the rest of the document. This manuscript has to be examined in toto on the basis of its internal patterns and the relationships of the tokens to each other. Anything else is guesswork for the simple reason that any sufficiently long collection of tokens will have a few that resolve into words and phrases in one language or another.





It's been my feeling for a very long time that if there is meaning in the text, it's a structured language, rather than a natural language, and the only reason I hold onto a shred of belief that it MIGHT be natural language is if the text is significantly manipulated (in terms of spaces and nulls, or a double conversion from a non-linguistic system that must then be further resolved into natural language).

But if it's not natural language, then what is it? I just don't think that something like "auto-copying" is a realistic option, if only because of the historical context. Something "real", some kind of information, must have been at the basis of this text.

It's not horror vacui or wishful thinking on my part. In fact I'd be perfectly fine with the text being just filler, because the images contain a ton of information for those who knew how to read them. They could stand on their own, or with just some labels. But the assumption that someone would generate this amount of nonsense text in this way "just because" seems very unlikely and unrealistic to me.

A structured language is invented. It requires a glossary or a grammar all its own.


For example (and this is just one quick example), Hildegard von Bingen set up her own system for creating words. It's years since I've looked at it, so I'm talking from memory, but I think she used "buz" or something similar as a suffix for plants/shrubs.

One might find an analog in the VMS -dy ending except that it is FAR too frequent to be anything in a natural language UNLESS it can expand out into a half dozen different optional endings (which immediately makes it different from von Bingen's system) and that creates a hornet's nest of possibilities and calls into question (once again) whether the spaces are real.

(26-05-2017, 08:14 PM)Koen Gh. Wrote: You are not allowed to view links. Register or Login to view.But if it's not natural language, then what is it? I just don't think that something like "auto-copying" is a realistic option, if only because of the historical context. Something "real", some kind of information, must have been at the basis of this text.

Why did you mean that the historical context speaks against the "auto-copying" hypothesis? To copy a book by hand was the only way to duplicate a book before Gutenberg invented the printing press. Therefore the "auto-copying" method fits very well into medieval times.

Your second argument is the same as Nick Pellings argument that the text is to complex to be meaningless. 

In a text using natural language the words depend on each other. You are able to determine word classes and you can found relations between this word classes. In other words if the text has some meaning you should be able to describe some grammatical rules and should be be able to find words which normally occur together. Only by using a good encryption method it is possible to hide this patterns for a text using natural language.

Until today nobody was able to determine different word classes for the VMS. It is also not possible to find repeated phrases as hints for some word order. It seems that a word in the VMS only depends on it's position in a line or page and on there position to similar words. For an encrypted text this would mean that the encryption method used was not only able to hide all the patterns typical for a meaning full text. The method was also able to generate some patterns on its own. This is only possible if the encryption method used is not very efficient since they is using fill glyphs and words. With other words the patterns found indicate at least that a word in the manuscript didn't stand for a plaintext word.

If the text is meaningless such problems didn't exist anymore. There is no need to hide meaning if there is no meaning in the first place. It is also no surprise if a text generation method used for a meaningless text would generate some typical patterns on its own. Features like that the words depend on their position and that similar words are related to each other are not only explainable with the "auto-copying" method the "auto-copying" hypothesis is based on this observations.

(26-05-2017, 05:22 PM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.Here is the statement:

"Cryptologists say the writing has all the characteristics of a real language."


Is that really true? Is that the general consensus among cryptologists? Is there a general consensus among cryptologists?

I'm very keen on the possibility that the Voynich text is ultimately a natural language or something very similar, and I wouldn't quite agree with this statement (not that I'm a cryptologist either).

(27-05-2017, 06:38 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(26-05-2017, 05:22 PM)-JKP- Wrote: You are not allowed to view links. Register or Login to view.Here is the statement:

"Cryptologists say the writing has all the characteristics of a real language."


Is that really true? Is that the general consensus among cryptologists? Is there a general consensus among cryptologists?

I'm very keen on the possibility that the Voynich text is ultimately a natural language or something very similar, and I wouldn't quite agree with this statement (not that I'm a cryptologist either).

This is what Cryptologists say about the Voynich manuscript:

"Languages simply do not behave in this way." ... "And yet I am not aware of any long repetitions of more than 2 or 3 words
in succession, as might be expected for instance in the text under the botanical drawings". (You are not allowed to view links. Register or Login to view. p.9)

"The short words, the many sequential repetitions, the rarity of one- or two letter words, the rarity of doublets (doubled letters), all militate
against simple substitution. So also does the strange lack for parallel context surrounding different occurrences of the same word as shown by words indexes. In the words of several researchers 'the text just doesn't act like natural language'." (You are not allowed to view links. Register or Login to view. p. 30)

The whole manuscript only consists of groups of glyphs which are similar to each other.  The combination-space for the glyph groups is restricted and the groups are self-similar in a given context like a page. This observation stands behind one of the ideas described by D'Imperio: After this idea the text contains words in a simple substitution cipher concealed in a longer dummy message. "The scribe, faced with the task of thinking up a large number of such dummy sequences, would  naturally tend to repeat parts of neighboring strings with various small changes and additions to fill out  the line until the next message-bearing word or phrase." (You are not allowed to view links. Register or Login to view. p. 31)

"That’s just the point — they’re not words!" (You are not allowed to view links. Register or Login to view.).

The manuscript contains only similar glyph groups whereas the ways to combine the glyphs to groups are restricted in some way. Yes, the groups are separated by spaces. In this way they look like words in a normal text. But they doesn't behave like words in a natural language: 
1) For instance it is not possible to distinguish different word types which can stand for nouns, verbs, adjectives or adverbs. 
2) It is not possible to identify some sort of grammar determining which words can follow each other. 
3) Longer sequences of repeated text are missing for the manuscript. 
4) The word length distribution is unusual for a natural language. 
...

The main question for the text of the Voynich manuscript is the way the glyph groups work. If you assume that the text contains meaning the question is what a group of symbols can stand for.