[split] Stolfi's Core-mantle-crust model

[MarcoP]

Am I the only one who finds it counter-intuitive to think of gallows as "core"? They feel much more fronted, I would probably put them in the front-part of the mantle. And then there is no core...

Hi Koen,
how would you parse the sample word qochepchody corresponding to the graph I posted?
The function of the mantle-core-mantle structure is to model how gallows (core) can be both preceded and followed by bench/e sequences (mantle). Of course, it is more frequent for the bench/e mantle to appear after the gallows as in 'qokeedy', but 'sheky' and 'cheky' total about 100 occurrences and they defintely look like legal words.
Counts for the various combinations can be found in You are not allowed to view links. Register or Login to view. (The Mantle Layer paragraph).

Words with both mantles are relatively rare, but frequent enough to look valid. The most common one might be 'chkeey' (13 occurrences, according to Voynichese.com).