The Naibbe cipher

[magnesium]

good morgning, a quick comment on the regularity of Voynich. all words that have a gallow sign integrated into the letter cth cph ckh cph also exist withouth the gallow sing in the middle. 
all words with an e, ee, eee, eeee in the word (and always and only after the first letter of a word) exist withouth the e in the manuscript. I question wether these theory of prefixes and suffixes could be used therefore? 

Thanks for your question! Within the Naibbe cipher, the prefixes and suffixes used to encrypt plaintext bigrams are generated using a modified version of Massimilano Zattera's "slot grammar": You are not allowed to view links. Register or Login to view.

The specific modifications I have made can be seen in Tables 8 and 9 of my preprint, but with a few exceptions, we're taking Zattera's slots 0-5 and using them to generate one set of affixes (what I call type-1 affixes; Table 8), while Zattera's slots 6-11 generate another set of affixes (type-2 affixes; Table 9). Most bigram prefixes within the Naibbe cipher are type-1 affixes, while most bigram suffixes are type-2 affixes.

We generate a given affix for the Naibbe cipher by going left to right within one of these tables and using 0 or 1 glyph per slot for all of the slots.

We don't have to generate type-1 affixes that contain gallows or bench-and-gallows glyphs, nor do we have to generate type-2 affixes that contain e glyphs. I simply mentioned those as just two of the examples for why this splitting of Zattera's slot grammar works well for this cipher: gallows glyphs never appear in type-2 affixes, and e glyphs never appear in type-1 affixes. So within the Naibbe cipher, these and other "affix-exclusive" glyphs can be handy ways of telling where the prefix-suffix breakpoint might be within a token that encrypts a plaintext bigram. But these glyphs are not the only ways of telling where that breakpoint is.