The Voynich Ninja
"Abnormal" words - Printable Version

+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Analysis of the text (https://www.voynich.ninja/forum-41.html)
+--- Thread: "Abnormal" words (/thread-4668.html)

Pages: 1 2 3

"Abnormal" words - stopsquark - 28-04-2025

While I'm aware that we don't know what the grammatical rules are for Voynichese, we have some general sense of how common certain character combinations are- for instance, a minim sequence is almost never found at the start of a word, consecutive benched gallows are fairly uncommon, etc. I'm curious about how many words we can find that apparently violate the conventions of the system.

One that I can think of off the top of my head- in the Rosettes map on You are not allowed to view links. Register or Login to view. , the center of the upper-leftmost path contains the word "ddssSex" in EVA. This is one of very few places in the MS that we see an x, a doubled s, OR a doubled d. What other rare word-forms like this exist? Bonus points if they're robust to transcription choices- i.e., if no matter how you interpret ambiguous letterforms, the words incorporate characters or strings that are uncommon.

RE: "Abnormal" words - oshfdk - 28-04-2025

Hi and welcome!

This can be automated, one can compute expected counts of bigrams given character frequencies and then compare with actual numbers. For example, I'm attaching a filtered transliteration file (based on #=IVTFF Eva- 1.7 with some simplifications) where some of the rarest bigrams made of common characters are highlighted. Then it's possible to take it one step further and look for rare bigram clusters, etc, etc. 


.txt   rarebigrams.txt (Size: 14.78 KB / Downloads: 28)

RE: "Abnormal" words - oshfdk - 28-04-2025

Interestingly, when filtering for lines that contain at least two instances of rare bigrams and relatively close to one another, I only get this:

Code:
% cat rarebigrams.txt | perl -ne "print if /\(.{,10}\(/"
f46v.8,=Pt      okeedchsy.qokeedy.chetedar.or.a( ic )k( yy )
fRos.14,@L0    ( dd )( ss )chx

So, it looks like the one from the Rosettes that you gave at the example could be the weirdest of them all. The other one, which is ai{cky}y in the transliteration file, is more like aicky or aiIky in the MS. Still quite weird looking.

RE: "Abnormal" words - Rafal - 05-05-2025

I have made a very similar thread not so far ago:
You are not allowed to view links. Register or Login to view.

I think most of us agree that there are "typical" and "weird" words in VM and weirds words are not so rare, possibly about 5-10% of words.
But what next?  Wink

RE: "Abnormal" words - Mark Knowles - 10-01-2026

The document below took me some time to produce:

You are not allowed to view links. Register or Login to view.

I have plenty to say about it, but I thought I would start by sharing it.

RE: "Abnormal" words - Mark Knowles - 10-01-2026

(10-01-2026, 04:44 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.The document below took me some time to produce:

You are not allowed to view links. Register or Login to view.

I have plenty to say about it, but I thought I would start by sharing it.

I have been going through the whole Voynich manuscript with the aim of compiling a list of the words with the most distinctive spellings. I have done this by crossing out in red on each page the most repetitive words as they seem to be and putting a yellow box around words which seem to me to have more distinctive spellings. I already had enough familiarity with the Voynich to be able to recognise the most repetitive words. Then of those more distinctive words I have put a green star next to those words which seem to be the most distinctive. Then I have copied those green starred words into a separate document, so I can see them altogether; this is the document above. Obviously to include all the annotated pages would leave a rather large file, though I can share those if desired. I have done this process manually for a variety of reasons. I don't know how accurate the Voynich transcriptions are especially when dealing with rare characters and where I might disagree with the interpretation of spaces.
 
I could have tried to automate the process considering Levenshtein distance or some other measure of distinctiveness, but I was not sure that I would be satisfied with the end result, although it could save time. Also, this process has forced me to look at the whole text and think about it.
 
So, of course, the end result is somewhat subjective.
 
I have covered all the pages. There are a number of pages with text written around a circle which is a bit of a nuisance to study. However I have managed to document them. I have ignored marginalia and other non-Voynichese text such as michitonese.
 
I have listed where each text appears on each page.
 
This is my first draft.

RE: "Abnormal" words - Eiríkur - 10-01-2026

Thanks, Mark. It's very interesting. It's a sort of a lightning tour of the document, too. I saw different forms of glyphs that subjectively seemed to support the multi-scribe hypothesis.

RE: "Abnormal" words - Mark Knowles - 10-01-2026

Well, I could give examples of the common Voynichese words, but people are often familiar with them and they are discussed a lot.

I have plenty more to say on this and will write more soon.

RE: "Abnormal" words - Jorge_Stolfi - 11-01-2026

(10-01-2026, 04:44 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.The document below took me some time to produce:

You are not allowed to view links. Register or Login to view.

I have been going through the whole Voynich manuscript with the aim of compiling a list of the words with the most distinctive spellings.

Thanks Mark for the work.  However, many of those "distinctive spellings" seem to be two more or less common words run together by the Scribe.  

For example, on page 014 (f6v), koarysar is probably koary.sar . The latter is fairly common; the former occurs only there, but the similar toary is fairly common too.

On page 010 (f4v), iiincheod was probably ain.cheod. Both words that occur multiple times. My reading is that the Scribe wrote the a slightly disconnected as two strokes ei (a glitch that he seems to have made dozens of times), and then the Retracer who restored the text turned that e into an i.  You can see the original e stroke, very faint, under the first i.

Another source of "unusual spellings" may be the prefixes o and y that are sometimes attached to the word, sometimes separated by a space of variable width. Sometimes the latter is encoded as ',' in the transcription files, sometimes as '.', and sometimes ignored...

On page 021 (f10r), the word must have been ro.tCho.Shor  or ro.tChy.Shor. All three words occur at least a few times.  The first strange glyphs was originally a normal r (you can see bits of the faint original ink) but was mangled into that weirdo by a Retracer.

Encrypted documents were usually written with great care, because even trivial errors, like replacing one glyph by a similar one, or skipping a glyph, would normally cause a non-trivial change in the decoded text.  Thus when one assumes that the VMS is encrypted one will necessarily assume that there are practically no errors.  

But there are several arguments for the thesis that it is not encrypted.  Then it most likely was transcribed from the Author's draft by a Scribe who knew the alphabet but could not read the text, who made many errors, which were not corrected because the Author could still read through them.  

Here is an example of the kind of errors that such an "ignorant scribe" would make:

   

The correct text should be 
  • Explicit dyalog9[=us] Bonavẽ[=en]turæ / inter animã[=am] et rationem
  • Here ends the dialogue of Bonaventura / between soul and reason
But the Red Scribe, who applied all the markings in red ink throughout the book, obviously did not know Latin, not even enough to tell that "Bonaventuræ" was a proper noun, that the "æ" was not "e", and that "raaonem" was not a Latin word.  And he misspelled the "Bona" as "Bone".  Thus the Proof Reader had to fix those errors (the brown ink).  

But the VMS apparently did not go through a Proof Reader...

And then you have the errors added by the Retracers who, not knowing the alphabet, often mangled the original faint glyphs as they retraced them.   Like that last example above.  (And this retracing is not just a fringe theory; you should take it as a fact, even if some desperately refuse to admit it...)

All the best, --stolfi

RE: "Abnormal" words - Mark Knowles - 11-01-2026

(11-01-2026, 12:54 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.
(10-01-2026, 04:44 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.The document below took me some time to produce:

You are not allowed to view links. Register or Login to view.

I have been going through the whole Voynich manuscript with the aim of compiling a list of the words with the most distinctive spellings.

Thanks Mark for the work.  However, many of those "distinctive spellings" seem to be two more or less common words run together by the Scribe.  

That may well be true in some instances. However, I would have to consider them on a case by case basis.

On page 014 (f6v) the spacing is unclear and so koarysar may well be koary.sar Although, I would hesitate from concluding two words were run together as the same could appear to be true amongst many words that we know in languages that we are familiar with. "super market" and "supermarket" are two distinct things with two distinct meanings.

On page 010 (f4v), your point about the first letter of the word originally being "a" is an interesting one.