Re: VMs: evidence that words are not independent of each other.

[Nick Pelling <nickpelling@xxxxxxxxxxxxxxx>]

Hi everyone,

At 12:31 18/08/2004 +0100, Marke Fincher wrote:
> For further evidence of neighbouring word dependence, consider the
> following word pairs which occur much more frequently in the
> forwards direction than in the reverse:
>
> (ar,aiin) f=22 r=3
> (ar,al) f=24 r=7
> (chol,daiin) f=34 r=7
> (ol,chedy) f=21 r=4
> (ol,shedy) f=21 r=7
> (or,aiin) f=52 r=6
> (s,aiin) f=29 r=3
> (shedy,qokedy) f=21 r=9

Note that most pairs on this list produce far more plausible "forward" than 
"reverse" Voynichese words - ie "araiin" looks good, but "aiinar" looks 
bad. This suggests to me that half-spaces might well have been transcribed 
as full-spaces in the transcription used here (which was it, BTW?) - EVA 
currently allows for no such distinction, though GC does use "-" for 
half-spaces in his.

Perhaps this same list of pairs might usefully be supplemented by the 
frequencies of the "forward" and "reverse" words produced by them?

Finally: as always, it's the anomalies that are most interesting - here, 
it's a good question as to why "ar al" should be 3.5x more frequent than 
"al ar". Perhaps these glyph pairs code for single letters in a Roman 
numeral kind of numbering system, obeying some kind of rigid adjacency / 
structure rules? It has long seemed curious to me that "ororor" in one 
place was broken up into "oro ror" to (it would seem) hide the otherwise 
obvious repetition of the pair. Perhaps a proper study of the structure of 
adjacent "or" "ol" "ar" "al" pairs in the VMs (but with all spaces removed) 
might be a good next step... Just as the Codex Seraphinus' numbering system 
was its weakest link, so might it prove to be the same for the VMs...

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


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