The Voynich Ninja

(23-04-2017, 10:14 AM)Torsten Wrote: You are not allowed to view links. Register or Login to view.Your approach is based on the idea that the VMS is random. Therefore my point is that VMS is not random or nearly random. If the VMS has meaning or not is indeed another question. 

I would not agree to your statement that different parts of the manuscript are generated slightly differently. In my eyes the opposite is true. The output of the generation method is changing over time. There is a transition from Currier A to Currier B. 
In the paper of Montemurro this is described as a network of relationships between the sections (see You are not allowed to view links. Register or Login to view.).
Herbal pages exists in Currrier A and in Currier B since after generating the pages in Currier A the pharmaceutical section the astronomical section were generated first.

Schinner has demonstrated in his You are not allowed to view links. Register or Login to view. from 2007 that the probability of the occurrence of a similar word decreases with distance: "Interpreting normal texts as bit sequences yields deviations of little significance from a true (uncorrelated) random walk. For the VMS, this only holds on a small scale of approximately the average line length; beyond positive correlation build up: the presence/absence of a symbol appears to increase/decrease the tendency towards another occurrence." (You are not allowed to view links. Register or Login to view.: p. 105).
There was a local element in the text generation method used for the VMS. At least the words are not independent from other words on the same paragraph. Therefore it is not possible to simulate the VMS with a purely randomly generated text resulting in words used independently from each other.

It's based on the evidence that it's random.   That of itself doesn't prove it's meaningless.   Any good cipher nowadays is random.   However, I'm highly skeptical of it having meaning as I can't imagine anyone back then encrypting it to make it look random, then applying a different process to make it look like an unknown language.

I don't think there's any conflict between our views that the Voynich Manuscript changes over time, as the manuscript was written, from Currier A to Currier B, or that my evidence in any way conflicts with that found by Montemurro and Zanette.

I don't have access to the full text, but Schinner's abstract states "The results significantly tighten the boundaries for possible interpretations; they suggest that the text has been generated by a stochastic process rather than by encoding or encryption of language."   This is what I've been saying.

(23-04-2017, 11:10 AM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2017, 02:07 AM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.

(22-04-2017, 11:06 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.I think it is well-accepted that the presence of words beginning with a gallows or ending with [m] is caused by line position, rather than their line position being caused by their presence.

Has anyone explained this, or provided any evidence?    Correlation doesn't necessarily imply causation.

This would be extremely easy to fake, whether or not the text has meaning.   Just ensure you start a new paragraph when you're about to write a gallows character.

Yes.

1) Words which begin with gallows at the beginning of paragraphs (Grove words) are often unusual in their structure. Removing the gallows makes them more regular. This suggests that the gallows is often added (although I agree that this is not the only cause).

2) If words ending [m] caused the end of a line, we would expect many short lines as words ending [m] occur elsewhere in the text, or even just shorter than average. But such lines typically run the full length of the page with no space for further words. This suggests that it is caused by being the last word in a line.

So what's unusual in their structure?   I only modelled paragraph breaks and line breaks very approximately as an afterthought. I begin new paragraphs on word-initial [p] or [f] only.   I know that isn't accurate ,and sometimes I shouldn't begin them there, and sometimes I should begin new paragraphs with words beggining with other glyphs.

I haven't looked in detail at the occurrence of word final [m] at all, but it's common enough that you can probably choose which [m]s to end the line at, to give the appearance of unbroken text.

I've attacked the problem by considering the VMS's large-scale properties, rather than homing in on specific weirdness, so there are plenty of specific things I haven't examined.   But I find it hard to believe, for example, that the meaning's hidden in the initial paragraph glyphs, or line-final glyphs, and the rest of the text is just filler.

And no one has been able to convince others that they have extracted any meaning from the text, or even identified the language.

(23-04-2017, 05:01 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.So what's unusual in their structure?  I only modelled paragraph breaks and line breaks very approximately as an afterthought. I begin new paragraphs on word-initial [p] or [f] only.   I know that isn't accurate ,and sometimes I shouldn't begin them there, and sometimes I should begin new paragraphs with words beggining with other glyphs.

I only modelled paragraph breaks and line breaks very approximately as an afterthought. I begin new paragraphs on word-initial [p] or [f] only.   I know that isn't accurate ,and sometimes I shouldn't begin them there, and sometimes I should begin new paragraphs with words beggining with other glyphs.

One obvious unusual feature is the presence of two gallows characters in a single word, which is rare throughout the text. Words with two gallows (in any order or combination) occur not more than 500 times. Yet over a fifth of those occurrences are paragraph initial: exactly where Grove words occur. Given that the average paragraph length is in excess of twenty words (often much in excess), this is an occurrence which is far too high to be random

Following your process, we could explain these statistics as a rule which starts a new paragraph when a word contains an initial gallows, especially if it contains a second gallows. Yet the simpler explanation is to explain both features by a different rule: add a gallows to the beginning of a word at the start of a paragraph. I don't actually think this is the exact rule, but I do think that it is a better one: both the initial gallows and the double gallows is explained by the single addition of a character, and paragraph starts are not determined by the arbitrary occurrence of a word structure.

Quote:I haven't looked in detail at the occurrence of word final [m] at all, but it's common enough that you can probably choose which [m]s to end the line at, to give the appearance of unbroken text.

I recommend you look at You are not allowed to view links. Register or Login to view. You can see that on some pages much more than half of all words ending [m] occur at the end of lines. Indeed, throughout the text about 75% of all such words occur in that position. Notably, You are not allowed to view links. Register or Login to view. has ten words ending [m], all of which are line end; You are not allowed to view links. Register or Login to view. has eight; and You are not allowed to view links. Register or Login to view. has six.

Quote:I've attacked the problem by considering the VMS's large-scale properties, rather than homing in on specific weirdness, so there are plenty of specific things I haven't examined.   But I find it hard to believe, for example, that the meaning's hidden in the initial paragraph glyphs, or line-final glyphs, and the rest of the text is just filler.

And no one has been able to convince others that they have extracted any meaning from the text, or even identified the language.

I, too, find it hard to believe that there are only patterns in a few places and meaningless filler in between. I think it more likely that we've not yet found all the patterns which are present. I've only been studying the Voynich text for about three years now, and I am constantly astounded by little twists and quirks which crop up in every batch of statistics I look at. My belief is that there any many layers to the text complicating any easy insight into its patterns.

Also, though I do believe the text is a natural language written plainly (more or less), I do not believe any claim to have discovered meaning or identified the underlying language. Nobody has convinced me that they've done better than guesswork.

(23-04-2017, 05:39 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.I recommend you look at You are not allowed to view links. Register or Login to view. You can see that on some pages much more than half of all words ending [m] occur at the end of lines. Indeed, throughout the text about 75% of all such words occur in that position. Notably, You are not allowed to view links. Register or Login to view. has ten words ending [m], all of which are line end; You are not allowed to view links. Register or Login to view. has eight; and You are not allowed to view links. Register or Login to view. has six.

Also, though I do believe the text is a natural language written plainly (more or less), I do not believe any claim to have discovered meaning or identified the underlying language. Nobody has convinced me that they've done better than guesswork.

I analysed the text without considering line breaks, so there's a chance I missed something important.    I'll examine the occurrence of word-final [m] when I have time.

If it's in an unencrypted natural language, it would have been identified and translated by now.

I think that the simplest theory that best fits the facts is that the text is generated by a stochastic process similar to the one I outlined in my blog.   If someone can come up with an alternative theory that fits the facts better, or is simpler, I'll be happy to concede.

(23-04-2017, 07:49 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.If it's in an unencrypted natural language, it would have been identified and translated by now.

This assertion (and any opposition to it!) is based on ignorance and untenable as a position. You can neither say "something is false because it is not known to be true", nor can I say "something is true because it is not known to be false". They are just generally bad arguments.

(23-04-2017, 08:08 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2017, 07:49 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.If it's in an unencrypted natural language, it would have been identified and translated by now.

This assertion (and any opposition to it!) is based on ignorance and untenable as a position. You can neither say "something is false because it is not known to be true", nor can I say "something is true because it is not known to be false". They are just generally bad arguments.

If there's no evidence for something, there's no reason to believe it.   Where's the evidence it's natural language?

I've provided evidence that the text of the Voynich Manuscript, ignoring the fine detail of its layout on the pages, was generated by a stochastic process.    I can reproduce text closely resembling the text of the real manuscript using the method I proposed, and have also shown that the word order, ignoring layout, of the Voynich Manuscript appears to be random (i.e. the probability of words is independent of the previous word, and is close to the value you'd expect by chance).

I suggest that the layout details are most simply explained as a postprocessing step.

(23-04-2017, 09:03 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2017, 08:08 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2017, 07:49 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.If it's in an unencrypted natural language, it would have been identified and translated by now.

This assertion (and any opposition to it!) is based on ignorance and untenable as a position. You can neither say "something is false because it is not known to be true", nor can I say "something is true because it is not known to be false". They are just generally bad arguments.

If there's no evidence for something, there's no reason to believe it.   Where's the evidence it's natural language?

I've provided evidence that the text of the Voynich Manuscript, ignoring the fine detail of its layout on the pages, was generated by a stochastic process.    I can reproduce text closely resembling the text of the real manuscript using the method I proposed, and have also shown that the word order, ignoring layout, of the Voynich Manuscript appears to be random (i.e. the probability of words is independent of the previous word, and is close to the value you'd expect by chance).

I suggest that the layout details are most simply explained as a postprocessing step.

You haven't provided evidence, even you don't believe that. You've simply provided a way to generate a very similar text. Just because you can make a text in that way does not mean the author did. This very same discussion was had with Rugg's theory. He beat you to the punch by over a decade, and I don't see what you've added to his theory.

(23-04-2017, 09:03 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.I've provided evidence that the text of the Voynich Manuscript, ignoring the fine detail of its layout on the pages, was generated by a stochastic process.  

I'm sorry but this is not correct.

You have generated a text using a stochastic process, and you argue that this is "similar" to the Voynich MS text.
It is equally possible to generate a meaningful text that looks just like the Voynich MS text. I explained that already You are not allowed to view links. Register or Login to view..
This demonstrates that your argument is insufficient.

"Similar" is just not sufficiently well defined.

Specifically, the only thing that's similar between your text and the Voynich MS is the word structure.
That is completely unrelated to the meaning of the text.

(23-04-2017, 09:26 PM)Emma May Smith Wrote: You are not allowed to view links. Register or Login to view.You haven't provided evidence, even you don't believe that. You've simply provided a way to generate a very similar text. Just because you can make a text in that way does not mean the author did. This very same discussion was had with Rugg's theory. He beat you to the punch by over a decade, and I don't see what you've added to his theory.

No, I have provided evidence.   One of the problems I have is that most people don't know what random looks like, and attach meaning to coincidences.

My theory is completely different from Gordon Rugg's.   Did you read and understand my blog articles?

(23-04-2017, 09:38 PM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.

(23-04-2017, 09:03 PM)DonaldFisk Wrote: You are not allowed to view links. Register or Login to view.I've provided evidence that the text of the Voynich Manuscript, ignoring the fine detail of its layout on the pages, was generated by a stochastic process.  

I'm sorry but this is not correct.

You have generated a text using a stochastic process, and you argue that this is "similar" to the Voynich MS text.
It is equally possible to generate a meaningful text that looks just like the Voynich MS text. I explained that already You are not allowed to view links. Register or Login to view..
This demonstrates that your argument is insufficient.

"Similar" is just not sufficiently well defined.

Specifically, the only thing that's similar between your text and the Voynich MS is the word structure.
That is completely unrelated to the meaning of the text.

Quote:An interesting test would be to check for repeating word sequences. This can be predicted with statistics, in principle, but also checked using software that some people have already available.

I carried out this test, and confirmed that words in the manuscript are independent of the previous word (see You are not allowed to view links. Register or Login to view.).   The mean frequency of word pairs was very close to the expected probability (the product of the probabilities of the two words considered separately), but the frequency's variance was very close to the square of the mean (i.e. a Gamma distribution, in my generated text it was a Poisson distribution), suggesting that the mechanism I initially suggested for deciding on transition paths, and only that, was wrong.   As far as I know, no one had spotted this before.