The 'Chinese' Theory: For and Against

[Jorge_Stolfi]

Could you please explain your theory in a few sentences (without using abbreviations) and give a very clear and concise example for laypeople like me?

Sure.  

The "Chinese" Origin theory (COT) proposes that the VMS is a phonetic transcription of one or more books originally written in a monosyllabic language, such as Chinese, Tibetan, Vietnamese, and several other languages of East Asia.  

A possible explanation of how that could have happened is that the Author was an European traveler who spent several years in some country where such a language was spoken (call it "China" for convenience). He saw that they had thousands of large books filled with knowledge totally unknown in the West.   He determined to bring some of that knowledge to Europe.

So he asked local doctors and scholars to point out the most important books on each subject. But he could not read the Chinese script, and had no hope of learning it. Thus he could not simply buy a copy of the book, or copy it himself.  Instead, he recruited some local Dictator to read each of those books aloud, while he wrote it down in a phonetic script or shorthand that he devised for that purpose.  Later, after he returned to Europe, he recruited a local Scribe to make a clean copy of his notes on vellum.

There are several other possible scenarios that would end up with the same result.  But I think the one above is the most likely one.

AFAIK the COT was first proposed around 2000 by the French linguist Jacques Guy. But it has been mostly championed by me, with practically zero acceptance by the community.

I have several bits of evidence that support this theory. The first one was the observation that You are not allowed to view links. Register or Login to view. (number of word types with length k) was strikingly different from that of European languages, but surprisingly similar to that of East Asian monosyllabic languages.  The cause for that peculiar distribution shape is that a syllable in those languages consists of a small fixed number of slots, and each slot can be filled with one element from some small set (like an initial consonant, a glide vowel etc) or left empty.  And eventually I got a descriptive model for Voynichese words that has such a slot structure.

There are several other properties of the VMS text that are at least compatible with the COT, if not evidence for it, but incompatible with its main alternative -- the "European" Origin theory (EOT).  I can elaborate if you wish.

That evidence convinced me that the COT was correct, and made me give up on the VMS -- because I thought that any further progress would require knowledge of Asian languages that I did not have and had no hope of learning.

But I was wrong on that last conclusion. 

Some years ago -- I forgeot pecisely when -- I learned that the most ancient and revered Chinese materia medica (book listing the known medicines and their uses)  had 365 "recipes".  This number roughly matched the estimated number of stars and parags in the "Starred Paragraphs" section (SPS) of the VMS, aka "Quire 20" or "Recipes", before the loss of its four central pages (f109r,v and f110r,v).  The Chinese book was the Shennong Bencaojing (SBJ), or the Classic Materia Medica of the (mythical) Divine Farmer (Shennong), believed to have been written roughly around 300 BCE. So it was easy to guess that the SBJ could be the source of the SPS.

But at the time I did not have a good copy of the SBJ, and anyway the Linux editors and utilities could not process Chinese text.  I recall getting some Japanese version of the book and checking that the length of the recipes was about right.  

It was only a few months ago that I got a digital file of the Chinese SBJ, and I learned how to process it with python. I could check that the distribution of paragraph lengths (how many parags have each length k) of the SPS matched fairly well that of the SBJ, after adjusting for the loss of the four SPS pages and the more verbose encoding.  

And then I found that the longest parag of the SPS had seven occurrences of daiin or similar words spaced apart just like the seven occurrences of the character 主 in the longest SBJ recipe. That recipe is about the meat and several other parts of the "red rooster" (literally "red male chicken").  In my view, that is pretty strong proof that the SPS is indeed a transcription of the SBJ.  

Since then I have checked 主 <-> daiin correspondence in several other SBJ recipes and SPS parags.  And I have identified another 2-3 potential "cribs", including 气 = "qi" or "vital energy", 血 = "blood", and 令 = "causes" or "makes"; but I am not yet totally sure about them.  More should start coming quickly now.

I still don't know the language, nor the encoding; except that each Chinese character does seem to correspond to about one Voynichese word, or five EVA characters, on the average.  That does not imply that the language is Chinese, since the Shennong Bencao was used in other countries too, where it was still written with Chinese characters but read in some local version of pseudo-Chinese.

On the other hand, the language and spelling seem to be the basically same through the whole VMS.  The differences seen between sections and between Herbal-A and Herbal-B may be due to the different subject, or maybe to some "spelling reform" that the author did to his system at some point.  Thus I conjecture that the other sections too are transcriptions of other classical Chinese books on the respective subject areas.

All the best, --stolfi