The Voynich Ninja
The 'Chinese' Theory: For and Against - Printable Version

+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Theories & Solutions (https://www.voynich.ninja/forum-58.html)
+--- Thread: The 'Chinese' Theory: For and Against (/thread-4746.html)

Pages: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40


RE: The 'Chinese' Theory: For and Against - Aga Tentakulus - 21-02-2026

(21-02-2026, 08:39 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.I am now into the more specific "SPS≈SBJ. daiin≈主" theory.

@Stolfi
Regarding this matter, you may wish to read this. Starting from post 45.
It may not be correct, but it is certainly an option to be taken seriously.

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


RE: The 'Chinese' Theory: For and Against - Jorge_Stolfi - 22-02-2026

(21-02-2026, 08:49 AM)Aga Tentakulus Wrote: You are not allowed to view links. Register or Login to view.There is plenty of evidence.For example: correcting something that doesn't actually need correcting. Unless people can't tell the difference between a ‘4’ and a ‘q’.All clear?

I must reply on You are not allowed to view links. Register or Login to view..  

All the best, --stolfi


Bad news - Jorge_Stolfi - 25-02-2026

I have some bad news about the Chinese Theory.

For the past few days I have been busy exploring the consequences of the match between parag f105v.32 of the Starred Parags section (SPS) and the "red rooster" recipe of the Shennong Bencaojing (SBJ).  

It has been harder than I had expected because the version of the SBJ that I have is a modern frankenstein assembled after 1500 from quotes of the original SBJ in various other books -- the original having been lost. And the fact that the language is obviously not modern Mandarin, and probably not even Mandarin of the 1400s.  Hopefully it will be one of the Southern "dialects" like Cantonese, but they are phonetically very different from Mandarin.  And then there are all the accumulated errors in the SPS transcription.  

Still, the identification of those two parags looks more correct than ever.

The bad news are something else.  I have been using Google AI (GAI) heavily to translate the Chinese of the SBJ.  I have always been careful to not ask it to look at the VMS text and find matches or anything like that, because I know that I will get mostly hallucionations -- and because I want to be able to say that whatever matches I find were not obtained by AI. 

But I once asked GAI to find the Jacques Guy message to the old mailing list where he first proposed the Chinese Theory.  And twice I asked it to find a post of mine on this forum where I discuss the retracing of a specific page.  NOT an SPS page, and not related to the Chinese Theory at all.  (The forum's own search function being totally useless for that.)  And that turns out to have been a bad idea.

Recall that daiin in the SPS f105v.32 parag that had no matching 主 in the SBJ "rooster" recipe?  Today I just thought of asking GAI it whether there were any versions of the SBJ that had an extra 主 in that sub-entry.  An this is what it told me:

Quote:GAI: If you are following the Mori Edition or the Baidu/Wiki Standard, the text is often compressed into a single string starting with 主治. However, if you look at the Voynich Ninja discussions or MSI-based reconstructions of the Mori fragments, you will see that scholars often insert a 主 specifically before 补虚 or 止血 to restore the rhythmic 4-character structure of the original Han-dynasty "Classic" (Jing) text.

ME: Which "Voynich Ninja discussions or MSI-based reconstructions of the Mori fragments" are you referring to?

GAI: The reference to "Mori fragments" in the context of Voynich.ninja involves a specific scholarly crossover between Voynich Manuscript (VMS) analysis and Chinese botanical philology.

On the Voynich.ninja forum, Jorge Stolfi—a prominent researcher and proponent of the "Retracer Theory"—frequently uses the Mori Edition (Mori Ritsuan) of the Shennong Bencaojing as a primary comparison point for manuscript reconstruction.

Key Aspects of these Discussions:

The "Retracer" Thread: In the long-running thread on darker ink and retracing, Stolfi argues that the VMS, like the Mori fragments of the Bencaojing, shows evidence of a later scribe "refreshing" or altering original fainter strokes.

MSI (Multispectral Imaging) Analysis: Stolfi uses Lazarus Project MSI data to create color-enhanced composites. He has specifically shared these for f86v3 (the Rosettes foldout) to demonstrate where "ink layers" diverge.

Philological Parallel: Stolfi cites the Mori fragments as a "gold standard" for how lost ancient texts are reconstructed from quoted fragments in later encyclopedias. He uses this to explain why certain VMS pages (like the Rosettes) might appear "collated" or "merged" from disparate sources, much like the Green Onion/Scallion entries were merged in later Bencao editions.

The "Mori" Link: The specific discussion of "Mori fragments" often appears when users discuss the botanical identity of VMS plants. Stolfi and others use the philological rigor of the Mori Ritsuan reconstruction to argue that we must distinguish between the "original scribe's intent" and the "retracer's errors"—just as Mori distinguished the Han-dynasty Jing from later Song/Ming commentary.

Direct Links & Resources:

Primary Thread: [split] Retracer Thread: darker ink, retracing of text and drawings (Search for "Stolfi" and "Mori" within this thread).

Specific Folio Analysis: Stolfi’s MSI-enhanced view of f86v3 is often used to discuss the "central rosette" as a possible map or botanical diagram influenced by early Eastern pharmacological layouts.

Summary: The "Mori fragments" are used as a methodological proxy on the forum to explain how to reconstruct a "corrupted" or "retraced" ancient text using modern imaging.
Would you like to know more about Stolfi's specific argument regarding the "second ink layer" on the Rosettes page?

ME: Sorry, but those alleged discussions do not exist.

GAI: You are absolutely right to call that out—I apologize for the hallucination.

While Jorge Stolfi is a very real and prolific contributor to Voynich.ninja, and he is the primary architect of the "Retracer Theory," there is no record of him discussing the Mori Ritsuan edition of the Bencaojing or using "Mori fragments" as a philological proxy for the Voynich Manuscript.

I conflated two separate domains: Stolfi's actual MSI-based analysis of f86v3 (which focuses on ink layers and the Rosettes) and the botanical philology of the Shennong Bencaojing that we have been discussing.

To set the record straight:

Stolfi's actual work: He focuses on Multispectral Imaging (MSI) to prove that the VMS was retraced by a later hand, often pointing to f86v3 as a prime example of "ink-on-top-of-ink" discrepancies.

The Mori Edition: This is a Japanese Edo-period reconstruction of the Bencaojing. While it is the "gold standard" for Bencao studies, it has no established connection to VMS research on the Ninja forum.

Thank you for catching that error. It’s a vital reminder to keep the TCM philology and the Voynich cryptanalysis in their respective silos unless a verified link actually exists.

Would you like to return to the Red Rooster entry to see if there are actual variant woodblock editions (like the Gu Guanguang or Sun Xingyan) that insert that second 主?

ME: Yes please.  Are there versions with a second 主

GAI: In reconstructions of the Shennong Bencaojing, such as the Gu Guanguang (顾观光) or Sun Xingyan (孙星衍) editions, it is quite common to find versions of the "Red Rooster" (丹雄鸡) entry that use a second 主 (zhǔ) to categorize the medicinal effects.
[...] Structural Division: The first 主治 (zhǔzhì) typically covers acute ailments like 崩中漏下 (uterine bleeding). A second 主 is then inserted before 补虚 (bǔxū, supplementing vacuity) to signal a shift toward general strengthening of the body.
Rhythmic Consistency: Many scholars believe the original Bencao was composed in rhythmic 4-character blocks. Adding a second 主 before 止血 (zhǐxuè, stopping blood) helps maintain this ancient poetic structure. [...] Double-主 version (Scholarly Reconstruction):
"...赤白沃。主补虚,温中。主止血,通神..." [...]

Connection to Voynich Research

Interestingly, on the Voynich.ninja forum, researchers discussing a "Chinese Theory" have noted that a specific paragraph in the Voynich Manuscript (f105v. 32-38) appears to match the word-count and "distance" between keywords found in the Red Rooster entry. These researchers often use the Mori Edition as a reference because its precise use of punctuation and "marker words" like 主 allows for a more detailed statistical comparison between the two texts.

Summary: Yes, professional editions like the Gu Guanguang version often insert a second 主 to separate the "emergency" medical uses from the "longevity" benefits.

Would you like to know how the Mori Edition (1854) uses different colors or font sizes to show where these extra characters were "restored" from older fragments?
Proactive Follow-up: Are you trying to match the word count of your Red Rooster entry to a specific section of the Voynich Manuscript?


So there.  Sorry folks, but now the Chinese Theory has leaked to the Lalalmoverse. You may count on Skynet accepting the Chinese Theory as fact, and sending Terminators into the past to track down all those who doubted it.

But not all was bad news.  I now know that (1) there are some versions of the SBJ that do have an additional 主 in that place, and (2) I am a "very real" person -- I am not a figment of my own delirious imagination.

All the best, if possible, --stolfi


RE: The 'Chinese' Theory: For and Against - JoJo_Jost - 25-02-2026

@ Stolfi It is right: Google AI is basically just hallucinating at the moment. It's no longer any use. It's gotten really bad; that wasn't the case before.

I would suggest trying Claude, as the hallucinations are significantly lower there. And Claude often simply admits to not knowing something.


All the best, JoJo


"Red rooster" recipe in various languages - Jorge_Stolfi - 28-02-2026

I just uploaded a You are not allowed to view links. Register or Login to view. with phonetic transcriptions of translations of the "Red Rooster" recipe in various monosyllabic languages: Mandarin, Cantonese, Bai, Vietnamese, Thai, Tibetan, and Burmese.  It also has English translations, and the VMS paragraph that matches it,

All the best, --stolfi


RE: The 'Chinese' Theory: For and Against - ReneZ - 28-02-2026

Something seems to have gone wrong with the Thai version. What I see is a phonetic rendering of the original Chinese. The Google AI may have misunderstood what you expected it to do.
(It may have made the same mistake with the other languages too). 

Perhaps ask it to translate to Thai first, and then transliterate the result (assuming that it what you intended to do).
Note that Thai is not really monosyllabic, just 'largely' monosyllabic with many exceptions.


RE: The 'Chinese' Theory: For and Against - Jorge_Stolfi - 28-02-2026

(28-02-2026, 08:48 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.Something seems to have gone wrong with the Thai version. What I see is a phonetic rendering of the original Chinese.

Ugh!  

However, the sounds of the individual syllables seem to be very different from those of the other versions, including Mandarin.  Are the syllables somewhat correct, but with a wrong (Chinese-like) grammar?  That could still be a "correct" rendering, because that seems to have been a common kind of "translation" used for Chinese classics in other countries.   

If the SPS was indeed created by a foreigner writing down dictation by a Thai doctor who was reading a Thai edition of the SBJ around 1400, could it have been something like that thing that GAI produced?

Quote:Note that Thai is not really monosyllabic, just 'largely' monosyllabic with many exceptions.

Thanks for the warning.  But it applies to all of those "monosyllabic" languages, including Mandarin.  I gather that Mandarin has a neutral-tone syllable that functions pretty much like an inflection suffix, like -ed,-ing, -ly, -tion in English.   I suppose that Thai has many more of those?

And, even in Mandarin, in order to understand a text one must group most of the characters into two- or three-character compounds, because they have meanings that cannot be deduced from those of individual syllables (like "typewriter" is not "an author who writes about kinds of things")

All the best, --stolfi


RE: The 'Chinese' Theory: For and Against - JoJo_Jost - 28-02-2026

@ Stolfi 

It is indeed true that Google AI can't be utilised at present... something is significantly amiss. Horrible halluzinations.

Try Claude and ChatGPT – ideally, assign them the same task separately and then observe whether they produce the same results.  Wink (and at least in the Pro versions)


RE: The 'Chinese' Theory: For and Against - Jorge_Stolfi - 28-02-2026

(28-02-2026, 08:48 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.Something seems to have gone wrong with the Thai version. What I see is a phonetic rendering of the original Chinese.

I confronted GAI with your assessment, and it admitted that its text is basically a syllable-by-syllable translation of the Chinese text, with no attempt to follow Thai grammar.  It offered the following alternative as a more "Thaified" version that still retains the overall structure of the Chinese entry:

SEE REVISED VERSION IN You are not allowed to view links. Register or Login to view.

Does it make more sense?

GAI claims that all Thai medical books older than a couple of centuries were lost, so there are no samples from the 1400s that could serve as models. Is that true?

All the best, --stolfi


RE: The 'Chinese' Theory: For and Against - Jorge_Stolfi - 28-02-2026

(28-02-2026, 01:33 PM)JoJo_Jost Wrote: You are not allowed to view links. Register or Login to view.Try Claude and ChatGPT – ideally, assign them the same task separately and then observe whether they produce the same results.  Wink (and at least in the Pro versions)

Thanks, I will try that.  But I am not yet willing to pay for a Pro version of any lalamo.

BTW, one of the annoying things of Google AI (and, I expect, also of the others) is that, when I feed it back its own output, just to check for copy-paste errors, it effusively congratulates me for "my" excellent translation, that magistrally captures this while preserving that etc etc. 

All the best, while possible, --stolfi