A new article published today by Celler Presse compares two independently developed approaches to the structure of the Voynich Manuscript.
Tobias Franke's "Functional Decoding Model" approaches the manuscript from a semantic-functional perspective, while my experimental model approaches it from the opposite direction: as a rule-based mechanical generation system, without assuming a translation or semantic meaning.
The comparison examines whether structures identified independently by the two approaches correspond to each other, including recurring word families, positional patterns and permitted transitions.
My model currently uses 19 states and 87 permitted transitions. In page-wise cross-validation it covered 92.84% of observed Voynich transitions, compared with 67.63% for a Latin control corpus under the same procedure.
I do not consider this a decipherment or proof of a historical mechanism. The purpose of the model is to formulate a testable and falsifiable hypothesis about how some of the manuscript's textual regularities might have been generated.

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

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

I would especially welcome criticism of the methodology and suggestions for independent tests.

Hello, everyone. i’ve registered to share my idea, because I didn’t find anything exactly similar on the site.

What if the VM is actually a medieval design guide on how to draw plausible illustrations for fantasy manuscripts?
As such, it could explain the content of botanical and astronomy sections.
Also it could explain its language, because it’s not a plain text but mostly a set of instructions how to produce illustrations from the same page.
I’ve asked ChatGPT and it found me the similar example of the same age: Göttingen Model Book (c. 1440). It is a book about generative drawing of ornaments.

Then, there is more elaborated, but more speculative idea, that explains why it’s encrypted. So the idea is, that it’s an internal handbook of highly skilled individual, who has created the fake manuscripts with ancient mysteries. As far as i know the 15th century saw a great interest in ancient books and also there are examples of such manuscripts. And to produce consistent mysteries over time someone had to record his ideas somewhere, but he obviously didn’t want for someone to uncover him, therefore the text is encrypted. This also explains why it is so hard to decipher, because private notes would contain some abbreviations, missing words, private terminology etc.

This idea could be tested, i guess, as similar plant illustrations should have similar text nearby, but i’m not technically skilled for that.

I hope I didn’t come with something that was discussed long ago and then discarded.

I think rdy is weird because of the apparent rules around where it is allowed to be used on a page / within text. 

rdy (and to a lesser extent ldy) are almost only ever found ending lines, next to images, as labels or first word of page. 
More simply, it does not seem to hardly ever exist in bulk text. 

"So what?" 
I have been wondering for a couple of days if all these cases are, or are part of, a label. Also how did the writer know this rdy containing word would land where it did? - Was it written first? Careful planning? - Random chance?
And also if this marker could be relatively reliable at finding where text starts/ends in unclear circular writing. If so, are there more that could aid in this.

Of 23 matches for rdy (Some of these could be in 2 category's), 

2 are the first word of a page
   

9 are in very close proximity to an image
   

5 are the last word of a line
   

4 are labels
   

1 could be "bulk text" or a "long label(?)"
   

1 could be "bulk text" or "line end" or "first word of a page", there's no indication on where the text starts/ends on this page.
It is found at "12 o'clock" on the page, so it's possible it's where the text starts or ends.
   

1 is certainly "bulk text"
   

So which ever way we slice the cake certainly 80%+ of cases are not in bulk text with 90%+ being possible.

Hi,

I have come across a new video about VMS.
It seems to me it is a nice intro video, but as I am about as literate in Voynichese as my dog, I will just post it so that others can check it.

It is presented by You are not allowed to view links. Register or Login to view. (podcast Host for Hidden History)

The Voynich Manuscript: This 600-Year-Old Book Nobody Can Translate
Hidden History and Today I Learned Science


Best regards,

I'm 95% convinced it's generated, but that might only be the top layer.. I have a feeling that there's plain text underneath that was encoded and that was then normalised through a generative layer, and all the tests that are run on the stats are only reflective of that layer which is why no one can quite get all the way to the bottom.. but given that doesn't reflect a single working practice I can find in 15th Century Germany, it may also be wrong...

I find Choe weird as e* (e followed by anything) happens 14263 times (Voynichese.com), egallows accounts for 833 - 5.84%

1. <f1v.7,+P0> choees - This one is actually "chores", discounted it from %
2. <f8r.6,+P0> choety
3. <f8r.9,*P0> tchoep
4. <f18v.5,+P0> ychoees
5. <f20r.6,+P0> choees
6. <f29r.7,+P0> choety
7. <f42r.12,+P0> choeky
8. <f43r.5,+P0> choetchy
9. <f54r.12,+P0> choek
10. <f66r.55,+P0> choekeey
11. <f67r2.49,+Pb> choer
12 . <f67r2.50,+Pb> choeea
13.  <f68v3.5,@Cc> choekcheey
14. <f68v3.14,@Cc> choekey
15. <f69r.3,+P0> choetey
16. <f70r1.2,@Cc> choekeeos
17. <f70r2.16,+Cc> choeky
18. <f70r2.17,+Cc> choetchaldy
19. <f70v2.1,@Cc> choeteedy
20. <f72v3.1,@Cc> choepchey
21. <f87v.7,+P0> qotechoep
22. <f88v.9,+P0> choeey
23. <f89v2.1,@Lc> choeesy
24. <f89v2.18,+P0> choeeer
25. <f106v.20,+P0> choefy
26. <f116r.15,+P0> pchoetal

When e is proceeded by cho the frequency of a gallows glyph following e jumps from 5.84% to 72%

It is also strange that 3 of 25 examples are double gallows words, which are extremely rare.

I spotted some faces on roots on folio f89r1 which I do not see on any other drawing of a plant/root.
I think, because it does not occur in other drawings (as I can see), this could be considered marginalia.
What do you think?    

Several months ago, I was browsing a random manuscript, and suddenly something struck me: the numbers looked a lot like those used for the VM folio numbers. (That is, the ones written top right on every recto, NOT the quire numbers).

I didn't do anything with this yet, since I have no idea how it compares to other manuscripts. Probably this style of writing numbers is extremely common. But how common exactly? Are some features rare or typical for a certain time/region?

Now I just had an idea: why not make a friendly competition out of it? Everyone can recognize numerals in manuscripts, so it doesn't require special skills, just some patience and investigation. My own submission is far from perfect and can be beaten. (I told Marco about it when I found it, so there is a witness to my having selected a manuscript ;))

I prepared a reference image of the VM foliation, taking into account variations for each number. Notice especially the two types of "3", one of which has a remarkable spike on top and a flatter top stroke.
   
I also prepared a spreadsheet, you could use your own copy of it to compare the VM foliation to other numbers you find. See here: You are not allowed to view links. Register or Login to view.

Rules

1. A submitted series of numerals must be from the same manuscript.
   
2. Try to find a full series, at least 1 -> 9.
   
3. When you want to submit a series, just PM me, or post it here in [spoiler] tags. My reason for keeping submissions secret would be to avoid a premature focus on a single time and place.
   
4. If you happen to find a better series later, you can submit that as well - more data points.
   
5. Try to compose an image or share your numbers in a copy of the spreadsheet, so that I don't have to hunt through your manuscript to gather numbers from various folios.
   
6. Manuscripts will remain a secret until the end, but I will sometimes provide a "leaderboard" based on my subjective feeling. After a while we will come up with a good way to rate.
   
7. PLEASE find something that's better than mine - it would look silly if I win my own competition!
   
8. At the end, I will add everything to the public spreadsheet.
   
9. The winner gets to pick a special color for their row in the spreadsheet as a reward ;)
   

• strategies for finding numerals
  
• which variations are easy to find, which numbers are harder to match
  
• specific properties of numbers we should pay attention to
  
• ...
  

Yesterday, Harvard physicist Matthew Schwartz unveiled a very powerful LLM harness for quantitative science called You are not allowed to view links. Register or Login to view., as well as a whopping 36 papers he has authored or coauthored using the tool. Among them is Schwartz's take on how the VMS was written, which is that it's iteratively generated gibberish:

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

His abstract:

The Voynich manuscript is an illustrated book of the early fifteenth century in a script no one has been able to read. Since Rugg, and then Timm and Schinner, showed that text with many of its statistics can be produced mechanically, quantitative study has come to favor a writing procedure over a cipher or an unknown language. Here we make that picture precise with quantitative tools: information theory, efficient methods for mixture-model computation using exact arithmetic, and solvers validated on planted test cases. Readings and text generators become probability models, fitted on half a scribe’s pages and judged on the rest. In six candidate languages, readings by simple or verbose substitution, consonantal writing, abbreviation, codebooks and the Naibbe cipher all fail on withheld pages. Such failures exclude a reading only where the test detects disguised real text of the same kind, so far only for scripture. Knowing a word barely helps predict the next, less than in any comparison text, glossaries included, yet the edge glyphs of adjacent words are coupled as strongly as in a recipe collection. The scribes reused their words, occasionally copied a recent one, and coined others by letter habit; this keyless procedure reproduces nearly all summary statistics of both principal hands. No proposed generator, ours included, reproduces that coupling. Two weak traces of content remain: distinctive label words and a faint dependence of word choice on subject. The low cost of the writing and the genres the pictures imitate suggest a book made to be looked at, not read.

I thought I would share this.

There is this manuscript called Dresden M206
You are not allowed to view links. Register or Login to view.

It is a bit about fortune telling and a bit about magic.

It seems to contain a lot of charms or things similar to a charm on the last page of Voynich.
Like the Voynich charm they seem to contain a mix of real and made-up words.

Some interesting words that give "michiton oladabas" vibes and others:
   

There is a lot of them. Maybe you will be able to find something more?