The Voynich Ninja

(12-07-2026, 11:40 PM)eggyk Wrote: You are not allowed to view links. Register or Login to view.Yes, but I was actually also referring to the order that they appeared in, too. Are there any eastern works that contain all of the sections in a similar way to in the VMS?

I don't know.  But the proposed scenario is that the Author copied (recorded / transcribed / translated / transliterated / whatever) several separate books, not a single book.  That is the case of the Shennong Bencao: it is only a list of 365 remedies (mineral, vegetable, or animal), the ailments they treat and/or the benefits they provide, and a terse and useless indication of provenance ("mountain valleys", "marshes", etc). Here and there the list of uses mentions alchemical ones (like, "chicken eggs can be turned into amber" or "iron can transmute copper and gold")   It has no illustrations of its own.  It has no description of the plants, very sparse and terse information on mode of preparation, no information about dosage, and no words on anatomy, astronomy, astrology.  

However, the book where the SBJ was probably copied from was a large medical encyclopedia that included drawings of the plants and possibly other illustrations, interspersed with the text. (I haven't seen it, but I read somewhere that the entry for "iron", for example, had an illustration of smelting and forging.)

So it is possible that, before or after copying the SBJ, the Author tried to copy the illustrations.  But the number of plants in the Herbal sections (97) is rather small.  IIUC, that Chinese medical encyclopedia had several hundred illustrations.

(By the way, @You are not allowed to view links. Register or Login to view.'s great tool refers to Nick Pelling observation that the Alchemists' Herbal usually has around 98 plants.   Which suggests that the VMS Herbal may be derived from this book. Or maybe the other way around... )

Quote:Well.. I wonder if the SPS is also quantitatively similar to the bloodletting and recipe sections of that german manuscript..

That section has only 21 or 22 "recipes".  Extrapolating for the four missing pages, the Starred Parags section of the VMS had  between 330 and 400 parags. The Shennong Bencao has 365 entries (maybe a few more in the source encyclopedia, because some of the original entries were split in two or more separate entries.)

All the best, --stolfi

(13-07-2026, 02:49 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.

(12-07-2026, 11:55 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.The matches that I have found so far indicate that, on average, Voynichese uses approximately 5 EVA letters for each Chinese character.   The following pinyin text with numeric tones (which may be the sort of encoding that the Author may have devised, if he was particularly clever) has 39 words and 169 non-blank characters

The average is close, but the spread / standard deviation is a very different story.

The question was whether Voynichese is actually faster to write than a hypothetical alternative spelling using Latin letters, like that one.  When measured in EVA letters it is a bit more expensive, but what if we counted strokes instead?  I don't know.

Yes, that "5 EVA for 1 hanzi" is only average. I didn't try to measure the standard deviation yet.  

I got that average ratio by comparing the lengths of gaps between cribs in matching entries -- between the 主 in hanzi and between the matching daiin in the SPS.  In the "red rooster" recipe the former are 23, 3, 3, 7, 4, 8, 5 hanzi, the latter are 110 EVA letters (should be 115, error of -5), 17 (+2), 22 (+7), 34 (-1), 19 (-1), 45 (+5), 25 (0).   So in the third gap the ratio is actually ~7.2 EVA per hanzi. 

All the best, --stolfi

(13-07-2026, 03:46 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.The question was whether Voynichese is actually faster to write than a hypothetical alternative spelling using Latin letters, like that one

In terms?
Compared to the You are not allowed to view links. Register or Login to view. handwriting, Voynichese is faster and easier to write than Latin letters. Its letters require only two or three lines and strokes.
If you're talking about that, of course...

(13-07-2026, 03:46 PM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.Yes, that "5 EVA for 1 hanzi" is only average. I didn't try to measure the standard deviation yet.  

I did not compute it either, but looking at Pinyin text with numerical tone markers (as you showed), it is clear that the vast majority will be in the range of 3-6 characters.

In your original (1990's) word length analysis you compared Voynichese with Vietnamese, which is a different story again.

On the dictation side, unfortunately (for both 'sides') it seems impossible to guess how someone - a European - would have interpreted the sounds of Chinese.
Just as a modern example, it is remarkable (to me) how much trouble Chinese people have in pronouncing Thai, even though both languages seem to have largely the same inventory. And this problem begins with hearing.

(13-07-2026, 05:02 PM)ololololo Wrote: You are not allowed to view links. Register or Login to view.In terms?
Compared to the You are not allowed to view links. Register or Login to view. handwriting, Voynichese is faster and easier to write than Latin letters. Its letters require only two or three lines and strokes.

How? Could you elaborate on which latin letters take more than two or three strokes? 

c, e, i, l, o, s, can all reasonably be written in 1 stroke
a, b, d, f, g, h, n, p, r, t, u, x, y  can be written in 2 strokes.
k, m, q, w, z can be written in 3 or 4 strokes. 

Note that other than 'm', most of the 3-4 stroke letters are relatively rarer (or potentially non existing) letters, too. 

e,  i,  l,  o,  d,  s   can be reasonably written in 1 stroke. 
a,  ch,  r,  k,  n,  y,  m   can be written in 2.
p,  f,  q,  t   can be written in 3. (t can be 1 or 3 depending on how curved the loops are)

Not to mention it is naturally quicker and easier to write in a system you have learnt and used and practised your entire life.

(13-07-2026, 05:02 PM)ololololo Wrote: You are not allowed to view links. Register or Login to view.Voynichese is faster and easier to write than Latin letters. Its letters require only two or three lines and strokes.

One or two strokes, actually.

(14-07-2026, 01:07 AM)eggyk Wrote: You are not allowed to view links. Register or Login to view.Not to mention it is naturally quicker and easier to write in a system you have learnt and used and practised your entire life.

That is a good argument ... against all the other theories.  

Why did the Author invent a new script, instead of using Latin letters?  

If the text was a known language in a simple substitution cipher, one could say that he invented a new alphabet to make the encryption safer.  But it is not a simple substitution cipher; if it is cipher, it is one that put the NSA to shame.  Then using Latin letters instead of that original script would have not made it any easier to crack.

If the goal was to produce a fake "Bacon Cipher Manuscript" to defraud an Emperor, which would be more likely to succeed: one written in the Latin letters that Bacon used and which looked like they could be deciphered one day, or in a script never seen before, which looked like an alien language that no one would ever be able to read?  If the intention instead was to make it look like a book from "Egypt or some other Oriental parts", why would the forger blow the ruse away by drawing people with European dresses and hats, European Zodiac icons, European-looking castles, etc? 

In fact, the Chinese Origin Theory seems to be the only one that can offer a plausible explanation for why the book uses an original script, and why the script looks that way.  Namely, for the same reason that people have invented new alphabets: to make it easier to write down a language whose sounds would not easily map to the alphabet they used themselves... 

All the best, --stolfi

(14-07-2026, 01:07 AM)eggyk Wrote: You are not allowed to view links. Register or Login to view.Not to mention it is naturally quicker and easier to write in a system you have learnt and used and practised your entire life.

Surely the Author did not devise the script in the morning and set out to transcribe the SBJ on that same afternoon.  When he started on the latter, he must have been experimenting and possibly using the script for some time.

It does not take much practice to learn to write a new alphabet as fluently as any other.  When I began learning Japanese, after a few dozen lessons I was writing the hiragana and katakana syllabics (~50 glyphs each) almost as fast as the teacher would say them.  

And it would have been much easier if I had invented the alphabets myself.  On the other hand,it would have been much harder if had chosen to study Chinese instead...

All the best, --stolfi

(14-07-2026, 10:11 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.to make it easier to write down a language whose sounds would not easily map to the alphabet they used themselves...

Is it that difficult?
It is much easier to compare the phonetics of Chinese with regular Latin than with any other alphabet.
It can be assumed that the author was a perfectionist and strove to convey the phonetics of Chinese as accurately as possible, ignoring the similarities with Latin. But... this is again an assumption. In fact, this is the whole Chinese theory.

(14-07-2026, 10:22 AM)Jorge_Stolfi Wrote: You are not allowed to view links. Register or Login to view.On the other hand,it would have been much harder if had chosen to study Chinese instead...

This is not the judgement of people I know who learned both! In Japanese you need about 2000 characters, which is fewer than the 3000 it takes to read Standard Chinese, but the reading conventions for Japanese are much, much harder. Multiple readings exists on single characters in Standard, but there usually is a dominant reading that a person need to be practically familiar with. In Japanese the typical situation is there are more than one common reading, and you have to know the etymology and context to select the correct one. Korean mixed script is commensurately easier on the same grounds because there are no native readings of the characters, so there is less to remember and work out from context; my experience was that was a lot easier than Japanese. The kana and hangeul (as the case may be) are nicer on-ramps to the characters, and for that reason romanizations now feature in a lot of Chinese education, including in China, but Japanese is often judged the most intricate to read.

Whether or not that judgement is correct, it occurs to me the principle here might actually clear something up for me as I continue to work on my much longer response about readings. When you were studying Japanese, did you get the impression that kanji, in the context of a sentence, had multiple correct readings?