The 'Chinese' Theory: For and Against

[eggyk]

As promised, here is the outcome of asking my program to find the matching SPS parag to the SBJ recipe BLAR 蜂子 = "bee larva", with more cribs.

There are several hanzi in that recipe that I am testing as possible additional cribs, besides 主 ↔ daiin = "mainly for"; but the two that I am reasonably confident about are 
气 = "vital energy", with canonical translation Chedy
令 = "makes", with canonical translation Cheody, Cheydy, or the like
When asked to look for the cribs 主,令,主,气,主, my program still says that f114r.4 is the best match. Here is the matching plot:

In this pairing, all hanzi cribs are paired with their "canoncal" translations, except 令 that has only a Chdy instead of a Cheody near the expected place. 

The gaps in the recipe before, between,and after those cribs are 2,12, 12, 8, 3, and 2 hanzi.  Assuming the average of 5.338 EVA lertters per hanzi, the predicted gaps in the matching SPS parag would be 11, 64, 64, 43, 16, and 11 EVA letters.  The actual gaps in f114r.4 between the assigned translations are 13, 64, 63, 40, 17, and 15 EVA letters.  The errors relative to the predicted values are +2, 0, -1, -2, 0, and +4 EVA letters.

All the best, --stolfi

I would appreciate if this could be explained in more detail. Are you only counting distances between the ones that line up? 

As far as I can see, only 1 of the 7 'chedy's match anything, and only 1 of the 3 'chdy's match. Unless i'm completely misunderstanding, you could have picked any hanzi character and got the same match rate. To be extremely clear, how i'm understanding that plot is like this sketch: 

   

Claiming a match in such a case doesn't make the slightest bit of sense to me. If i've misunderstood something here please let me know as I don't intend to mis-state what you have shown.

Seperately:

If this is what you are doing and acknowledge that the other non-matching instances will need to be explained at some point, an interesting method could be a brute force one: calculate the average position (relatively within recipes) for each potential 'block' of voynichese for each parag. For 'blocks' that occur multiple times on average, find and plot the average position for the 1st, 2nd, 3rd... Nth instance within parags. 

This would effectively provide the answer to "If I were to pick an SPS parag at random, where would I expect these voynichese 'blocks' to most likely occur?". 

Then, you can do the same for each chinese character across the SBJ recipes and compare the data, mapping the most similar to eachother. If the data actually holds valuable information, it would be analogous to using letter frequencies to decode mono-alphabetic substitutions. 

Then, using the 'block' -> hanzi mappings, pick a parag, write out the matched chinese characters and see if the recipe matches any of the actual ones. Even a relatively low score would be encouraging in such a case. In fact, any recognisable part of any recipe would be encouraging. 

Just a thought, easier to say than do