Re: VMs: Artificial voynich - Its time to put up !!!!

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Dear Adam,

Initially, Jeff's approach used one table per EVA letter (which, given that 
EVA is a modern stroke transcription retrofitted to Voynichese, seemed 
somewhat likely to be wide of the mark).

More recently, he has refined this to add attributes to letters that only 
allow them to appear in certain positions (ie, at word-start, word-middle, 
or word-end).

This is, as you suspect, a method of generating VMs-like text using a 
pre-decided set of language-like statistics (in his case, those implicit in 
an (unnamed) Italian plaintext). The more rules and filters he applies, the 
more it will resemble Voynichese - but the pay-off per additional rule will 
rapidly diminish. Naturally, at no stage will it be Voynichese. :-(

Jeff also adds:-
	"It also has links to tag the same for triplets [by] having each
	pair link to a list of following letters, also marked with the
	flags for initiating a word, continuing a word or ending a word.
	It is quite a compact table."

This sounds like yet more adjacency rules to filter out bad-looking 
"sentences" (by hacking word-final-to-word-start transitions). So many 
rules, so little pay-off! :-(

Cheers, .....Nick Pelling.....


______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list