The Voynich Ninja

Anyone who consults this forum will see that, referring to the script, the terms text and word are used extensively in many of the threads. It is assumed that the union of glyphs forms a word and that glyphs are the constituents of the alphabet of a language, whether natural or encrypted.

  This is something that is far from being proven and is a way of speaking that has conditioned and still conditions the view of what this script is. The use of the terms word and text is not irrelevant, as it influences our perception. When discussing the combination of glyphs, it would be more accurate to speak of a group or set.

(30-12-2025, 10:52 AM)Antonio García Jiménez Wrote: You are not allowed to view links. Register or Login to view.The use of the terms word and text is not irrelevant, as it influences our perception.

That is true.  However, the word "word" can mean very different things depending on the language.  

The Chinese generally thought of each syllable (= each Chinese character) was a "word", until Westerners found that concepts that were one word for them usually mapped to two syllables in Chinese.  Then, when transcribing Chinese with Latin letters, they often would run those two syllables together.  As in "Beijing", "wushu", "wonton", "jiaozi"...  AFAIK that is the case also for Tibetan, Vietnamese, and other "monosyllabic" languages.

A compound term is two or more word that have a specific meaning that cannot be inferred from the separate components.  In many languages, some compound terms are often written as a single word, with or without hyphens: "typewriter", "windshield", "inkwell"...  while other compounds are written separately: "pit stop", "solar sail", "gas pump", ...

In Italian the postfix oblique pronouns and some modifiers are traditionally written attached to the verb: "portioamocela" = "let's take it there", "ditecelo" = "tell it to us".  In Spanish and Portuguese they may be hyphenated to the verb.  In other languages (like English) they are separate words.

The definite article "al-" in Arabic script is written attached to the noun. In Gaelic, IIUC, the definite article is a suffix attached to the word. In most Romance it is a separate word, but in Italian it is sometimes attached with an apostrophe: "l'uomo".

In Romance languages one usually turns an adjective into and adverb by inflecting it as feminine singular and attaching the suffix "-mente", "-ment", or similar; equivalent to English "-ly".  "lento" -> "lentamente".  However, phonetically that suffix behaves like a separate word, and should logically be written as such: "lenta mente" or "lenta-mente".

And don't ask about Turkish or Hungarian...

So, using the word "word" to mean "sequence of Voynichese glyphs separated by thin spaces and delimited by wider spaces" should not lead us astray, if we keep aware that "word" can mean different things in different languages.

Besides, the Voynichese "words" defined this way have properties that are shared with what one consider "words" in most languages.  They behave as atomic units with respect to line breaks (so far we have not found evidence of hyphenation in the VMS). They have a rather rigid structure, and labels are often a single "word" with that same structure.  The number and frequency of such "words" follow Zipf's law.  The entropy per word is within the range of natural languages with their traditional scripts.  And so on...

All the best, --stolfi

I see you're an active forum member and post in almost every thread. But it's ironic that you're giving a lesson in universal linguistics in this thread titled "No text, but visual code."

  You are a prime example of the failure to find a language in the Voynich manuscript. You have been trying for almost 30 years without any success. You have certainly done a good job in analyzing the script structure, but you have not been able to reach any conclusion that is accepted by the rest of the researchers.

   I remind you that there are excellent researchers who have not seen any language, such as Timm and Schinner, or Gaskell and Bowern in their paper Gibberisch after all?...

   You'll certainly never convince me. I have a pretty clear idea of what the Voynich script is.

(31-12-2025, 09:25 AM)Antonio García Jiménez Wrote: You are not allowed to view links. Register or Login to view.I see you're an active forum member and post in almost every thread. But it's ironic that you're giving a lesson in universal linguistics in this thread titled "No text, but visual code."

Posts in a thread must be about the subject of the thread. For or against. Not exclusively for it.  

When you make a claim "Anyone who consults this forum...", you must expect replies from all the Anyones who do not agree with it...

Quote:You are a prime example of the failure to find a language in the Voynich manuscript. You have been trying for almost 30 years without any success. You have certainly done a good job in analyzing the script structure, but you have not been able to reach any conclusion that is accepted by the rest of the researchers.

I did find a language: Voynichese.  Many here who have looked at the evidence seem to agree that it is indeed a language.  What they don't agree on is what kind or language it is, and/or how it is spelled/encoded.  But no one has found an answer that others will agree with.  That will come, eventually...

Quote:there are excellent researchers who have not seen any language, such as Timm and Schinner, or Gaskell and Bowern in their paper Gibberish after all?...

"Did not see any language" ind not the same as "Proved that it is not any language".

Those researchers  only found some statistical property of the VMS text that they claimed could not exist in any natural language, and concluded that it was gibberish.  (Actually Gaskell & Bowern stopped short of such conclusion.)

But those researchers failed to understand that statistics are properties of the text, not of the language.  If I were to write that it is not text but a visual code and furthermore it is a visual code not text and by visual code I mean not text no no not text but only a visual code and being a visual code it is not text and it is definitely a visual code and not just text where text means not a visual code which is visual not textual as in a text that is not visual, this text would have very bizarre statistical properties -- and yet it is plain English, as grammatically correct as it could be without punctuation.

Partly because of that mistake, they did not satisfactorily show that "natural languages" do not have those properties.  First, they compared the VMS only with samples of novel- or textbook-like prose, not herbal or pharmacological texts.  If you check the text of medieval herbals, like You are not allowed to view links. Register or Login to view., you will see a lot of repetitive patterns -- that would probably match the "setf-citation features" of Thorsten & Timm.

Second, those authors did not test their claim against a sufficiently broad sample of languages.  As I have pointed out since ~1998, the distribution of lengths of the VMS lexicon is quite unlike that of polysyllabic languages --- like all "European" languages -- but matches quite well that of monosyllabic languages.  And many others have noticed that the rigid structure of Voynichese "words" is quite compatible with them being syllables, rather than polysyllabic words.  But Gaskell & Bowern compared their VMS metrics only with those of texts in polysyllabic languages...

And finally none of those who defend the "meaningless gibberish" theory have found a method that could (a) produce text with statistics similar to Voynichese, (b) did not need large tables or rule sets specifically tuned to reproduce the "natural" statistics (like Zipf's law) that we see in Voynichese (which were unknown at the time!), and © would have been easy enough for the Author to use through all those 200+ pages.

As I explained before, it is mathematically impossible to prove that a string of letters (bits, sounds, etc.) is gibberish.  At best one could find a short algorithm in a simple programming language (like a Turing machine with a short input tape) that generates that precise sequence (not just a sequence with similar X, Y, and Z statistics).  That would prove that the "message" of the text is just that short algorithm.  Otherwise there always remains the possibility that the text contains is a non-trivial meaningful message, just encoded in a "random-looking" way.

Quote:I have a pretty clear idea of what the Voynich script is.

IIRC, in a recent post you said that you think the "message" is encoded not in the sequence of glyphs as we transcribe them, but in subtle details of how the strokes of those glyphs connect (or not) to each other.  Is that a fair summary of your theory?

That would be a form of steganography, a general encryption method that Trithemius described in the 1499 book Steganographia (whose third encrypted volume was finally cracked in 1998 by "our" Jim Reeds!).  The specific form that you seem to propose was proposed sometime around 2000 to  the mailing list.  The main proponent (whose alias may have been Rayman Maleki and/or Glen Claston, not sure) hated EVA and devised his own transcription alphabet that would record all those subtle details. 

But he could not crack the VMS either, and apparently did not make many converts. (IIUC his transcription was eventually included by Rene in the IVMT file.)

Anyway, given that theory, I understand why you hate my theories about the VMS -- not just the "Chinese" Origin Theory (COT), but also the Ignorant Scribe Theory (IST, that says that the Scribe copied the Author's draft after learning the alphabet and nothing else), the Sparse Retracing Theory (SRT, that says that many glyphs were retraced by later owners) as well as the Massive Retracing Theory (MRT, that says that in fact almost all the text was retraced at some point).  

Well, I can't do anything about that.  For now...

All the best --stolfi

To begin with, I don't hate your theories. I simply reject them because I don't find them well-founded.

You can write in this thread and refute my ideas, but please first read carefully what I think, because I don't recognize myself in the researchers who, according to you, had or have similar ideas to mine. As far as I know, no one has proposed anything similar to what I am proposing, which is perhaps why this thread is generating curiosity. To put it in today's terms: I think outside the box.

  The computational and statistical analyses that you and many others like you do have some use, but they forget something essential: that we are dealing with a medieval document made by people who thought differently from us, and therefore there is an unpredictable factor that escapes any conventional analysis.

In your view of the Voynich Manuscript, as in that of others, the images are secondary and hardly merit study. For me, however, the images are an essential part of the codex, and may even explain why the script is nothing more than a more or less ordered set of astronomical symbols.

(31-12-2025, 04:58 PM)Antonio García Jiménez Wrote: You are not allowed to view links. Register or Login to view.You can write in this thread and refute my ideas, but please first read carefully what I think, because I don't recognize myself in the researchers who, according to you, had or have similar ideas to mine. As far as I know, no one has proposed anything similar to what I am proposing, which is perhaps why this thread is generating curiosity.

Please have pity of me: I just joined the forum this year, and this thread is already 1600+ posts long.  Could you please re-post a summary of your theory, or point to a post with that information?  Thank you...

All the best, --stolfi

Maybe another day.

I wish you and all forum members a happy new year.

One of the biggest problems in Voynich research, in my opinion, is that attention has been primarily directed to the script, while the images have been studied individually and almost never as a whole. It's curious because it's the opposite of what people in the Middle Ages would probably do. Illiterate people, of course, but also educated people if they had a codex full of fascinating images in front of them.

  I believe that the authors of the Voynich were primarily concerned with exhibiting the fruit of their powerful imagination and only later attached a mechanically produced script to accompany the imagery. 

If they had had something important to say in a text, they would have written it in Latin or their original language, but that was not the case because for them the script was only a means of adding more mystery to the wonderful images they drew.

I don't see it that way.
What used to be secret because it was newly discovered was a reason for encryption. Only for the initiated. How long did China keep the secret of black powder production a secret?
Today, you can get the instructions and ingredients in a chemistry set for children.

(02-01-2026, 12:56 PM)Antonio García Jiménez Wrote: You are not allowed to view links. Register or Login to view.One of the biggest problems in Voynich research, in my opinion, is that attention has been primarily directed to the script, while the images have been studied individually and almost never as a whole. It's curious because it's the opposite of what people in the Middle Ages would probably do. Illiterate people, of course, but also educated people if they had a codex full of fascinating images in front of them.

This is a very distorted view of Medieval manuscripts.  We may imagine them as being highly illustrated, but that is only because those books with nice pictures had a much better chance of surviving to this day, being exhibited in museums, being reproduced in books, magazines, websites. etc.  And being mentioned here.  

But I bet that more than 80% of all manuscripts written in the 1400s had practically no illustrations, You are not allowed to view links. Register or Login to view..

And that is also a very distorted view of Voynich research.  I bet that, since before Barschius, ten times as many man-hours have been spent poring at the figures than at the text.   Think of all the time that Petersen and others have spent trying to identify the plants, for instance.  Because there was little that people could do with the text, until it was transcribed into computer-edible form. 

And even now, in the computer-assisted era, most newbie Voynichologists quickly find that their statistical grinding does not lead anywhere, and give up...

All the best, --stolfi