The Voynich Ninja

(18-02-2016, 09:39 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.Without saying that this is the solution, but just to give an example and some food for thought:

if the words were all Roman numbers (i ... ix ... lvi ... cliii) 
the same feature would emerge. Each word would have several ones that differ from it by
only one addition, deletion, replacement or swap.

Like a code book... Another option is that a constructed language was used.

I accept both of these as plausible, though not highly probable given the inherent complexity of the encoding process.

Why is it that the VM's character entropy is relatively low, yet the word variability is relatively high? Normally, i would expect these to be mutually exclusive.

For example, if you take an english text and perform random insertions, removals and replacements, the result should be a text with both higher entropy and higher word variability because word structure and patterns are reduced.

In a sense, Voynichese has structured noise, where character insertions, removals and replacements are not random and instead follow a set of rules.

This could indicate the usage of null characters (e.g. 'a' after 'e' is null), but the word lengths don't really accommodate it...

-Job- Wrote:This could indicate the usage of null characters (e.g. 'a' after 'e' is null), but the word lengths don't really accommodate it...

Maybe the null characters could be used like the null character in hangeul, ie. only in the beginning of a syllable?

I think maybe the manuscript could be written in some form of Korean. It was dated to the 1430's, which is the decade before King Sejong promulgated hangeul (completed in 1443-44, published in 1446), so it falls exactly during the most likely period in which he would have been researching the new script he wanted to invent for Korean, and we also know Europeans traveled to the far east back then, so someone might have ended up in Korea, wanted to share some of his knowledge, learned what was then King Sejong's working version of the new Korean script, probably through a third source, and then wrote down his knowledge in the Korean language (probably with a native speaker dictating the Korean translation) in said script, only to then be essentially told it was all in vain, as King Sejong had scrapped that script already and was developing a new one. The European traveler then took the manuscript back with him to Europe, and then it eventually made its way to people who had no idea of its origins and hence it has been baffling us to this day.

The reason for that is some features mentioned in this very thread. First, the very high sound replacability, which exists in Korean too (replace any sound in a syllable with another from the same category, and you have another perfectly valid syllable), then the apparent ng-sounding chararacter that only appears at the end of words (though it should be look if that's not perhaps the end of syllables), as the only East Asian language at the time that had both high sound swappability and a ng sound that only appeared at the end of syllables, was Korean. Add to that that one or more null sound signs were proposed for the Voynich manuscript, and the a null sound sign is a feature of the writing system currently used for Korean, hangeul, nods us once more at Korean.

Now, just to preempt possibly accusations of being a Korean nationalist - I'm not, I'm not even Korean, and I don't even know the Korean language as much as I wish, I have just always thought of the manuscript as having ties to East Asia (and Stolfi and partially Diane O'Donovan seem to think that way too).

Also, I do note that Korean has a whole lot of vowels, while it seems only three have been identified for the Voynich Manuscript, two of which seem to be variants of each other - this is IMHO not incompatible with Korean, which while as I said before, it has a whole lot of of vowels, they are all written in Hangeul with combinations of three elements - ㅡ ㆍ ㅣ. So the Voynich manuscript might have three vowel glyphs but a whole lot of actual vowel sounds written with combinations of them.

Now, my hunch might well be wrong, but who knows. :p

(19-02-2016, 07:02 AM)Battler Wrote: You are not allowed to view links. Register or Login to view.

-Job- Wrote:This could indicate the usage of null characters (e.g. 'a' after 'e' is null), but the word lengths don't really accommodate it...

...
I think maybe the manuscript could be written in some form of Korean.
...

In my opinion the VM has lots of elements consistent with a 15th century manuscript from central Europe (Germany in particular), including glyphs.

It's true that the text itself is not really compatible with European languages, but so far that's attributable to a general incompatibility with natural languages as a whole.

I think that will continue to be the case as we perform comparative analysis with additional languages.

If you have some sample Korean texts that you think are a good match for Voynichese, i'm willing to run some tests to confirm or dispute that.

- -Job-: Unforuntately, all we have is Korean documents saying King Sejong's protect was finished in 1443-44 and then King Sejong's book from 1446. We have nothing left of his actual research and earlier attempts to come up with a script, so we can only speculate.
Now I would suggest taking any Korean text from the period, preferably one written completely in hangeul, then rewrite it so the jamo (single components of hangeul syllable blocks) are written sequentially like an alphabet, and decompose all vowels into their constituent components too. In this way, eg. the world hangeul itself, written 한글, would be written ㅎㅣㆍㄴㄱㅡㄹ. Then run statistics on thus written Korean text and see if it matches Voynichese in any way.

Edit: And even if the script turns out to be Korean, the language itself itself might not necessarily be. Even hangeul had several now obsolete jamo that were designed to write down sounds found in contemporary Chinese but not Korean, so if this is indeed a preliminary pre-hangeul script of King Sejong, it might write down Chinese too, so the manuscript might contain Korean, Chinese, or even both which was common to write in Korea at the time, and which could explain the seeming two "languages" of the manuscript.

-job-

i am feeling stupid, because i do not understand your numbers on "replacement counts"

Quote:In Pliny's vocabulary, if we take a word containing an "m" and replace it with an "s", we get a valid word 7% of the time. There are other valid substitutions but they yield less than 5% - i'm only listing the most significant ones.

Bible (Latin):
m replaced by s: 136/1887 7.0%
o replaced by i: 107/1828 6.0%


"The following gives the percentage of words that are still valid after a character substitution, in some texts of roughly equivalent lengths"
Pliny:  m replaced by s: 213/3074 7.0%

What does this mean in plain English ?

What is the software you use for this analysis, what rules ?

(22-02-2016, 04:10 PM)Davidsch Wrote: You are not allowed to view links. Register or Login to view.-job-

i am feeling stupid, because i do not understand your numbers on "replacement counts"

Quote:In Pliny's vocabulary, if we take a word containing an "m" and replace it with an "s", we get a valid word 7% of the time. There are other valid substitutions but they yield less than 5% - i'm only listing the most significant ones.

Bible (Latin):
m replaced by s: 136/1887 7.0%
o replaced by i: 107/1828 6.0%


"The following gives the percentage of words that are still valid after a character substitution, in some texts of roughly equivalent lengths"
Pliny:  m replaced by s: 213/3074 7.0%

What does this mean in plain English ?

What is the software you use for this analysis, what rules ?

What don't you understand, specifically?

The portion you quoted indicates that, among the set of words used in the Latin Bible sample text that contain an 'm', replacing the 'm' with an 's' results in another word used in the same sample text about 7% of the time.

The software i used was my own, which has also generated You are not allowed to view links. Register or Login to view..

(24-02-2016, 07:35 AM)-Job- Wrote: You are not allowed to view links. Register or Login to view.The portion you quoted indicates that, among the set of words used in the Latin Bible sample text that contain an 'm', replacing the 'm' with an 's' results in another word used in the same sample text about 7% of the time.

Hello Job, I expect that almost all of the Latin substitutions depend on the last letter: 
e.g. 
factum (Accusative / Neutral) factus (Nominative / Masculine)
essem (First person) esses (Second person)

While in Voynichese, substitutions are not clearly related to the position of the character in the word. Is it so?

Well, who/what  defines what is still a readable word in general ?
And how on earth do you define that for the Voynese words ?

I tried such a thing before and have so many examples what can be replaced 
in a word and in a sentence, that almost ANYTHING is possible.

So probably I am missing something essential here, or anything can be replaced by anything, see below.
In the last case, we can generate many other numbers.

Your program that generated the "valid" replacements has one or more rules defined based on position or cohesion?
 or do you simply replace ANY given letter?


------
Just some examples:

The quick brown fox jumps over the lazy dog

Glue:
variation1: Thequick brownfox jumpsover thelazy dog
variation2: Thequickbrownfox jumpsoverthelazy dog
etc..

One or more letter replacement
1he quick brown fox jumps over 1he lazy dog
1he quic1 brown fox jumps over 1he lazy dog
1he quic1 brow1 fox jumps over 1he lazy dog
1he quic1 brow1 fo1 jump over 1he lazy dog
etc..

vowels:
Th1 qu1ck br1wn f1x j1mps 1v1r th1 l1zy d1g
etc. etc.

deletion/subtraction:
Th quck brwn fx jmps vr th lzy dg
variation1:
T_ quick brown fox jumps o_ t laz_ dog

random insertion:
The2 qu2ick br2own fox2 ju2mps ove2r 2the l2azy d2og
etc.

and many many more other possible variations

(24-02-2016, 11:47 AM)Davidsch Wrote: You are not allowed to view links. Register or Login to view.Well, who/what  defines what is still a readable word in general ?
And how on earth do you define that for the Voynese words ?

I tried such a thing before and have so many examples what can be replaced 
in a word and in a sentence, that almost ANYTHING is possible.

So probably I am missing something essential here, or anything can be replaced by anything, see below.
In the last case, we can generate many other numbers.

Your program that generated the "valid" replacements has one or more rules defined based on position or cohesion?
 or do you simply replace ANY given letter?

You seem to be completely misinterpreting the results i posted.

My analysis only shows that, relative to the sample texts used, the VM has higher character replaceability values.

How you account for this feature is up to you. It could be the result of extremely irregular spelling for example.