Re: VMs: Word structure, glyphs...

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi Matthew,

At 21:40 13/07/2003 +0100, Matthew Platts wrote:
> The rantings of a madman follow. If it's unstructured, I can excuse 
> myself: I've been listening to Chris Morris' Blue Jam...

FWIW, I found all his "doctor" riffs funniest. :-)

> I would tend to imagine what I would call a 'compound symbol' in the 
> Voynich language - sets of symbols which always exist together - to be a 
> discrete component of the language;

I've been on this "pairification" path for a while now, and the logical 
position it leads to (where everything is pairs --> the "interior language" 
is illusory --> everyone else is being led by their noses by the 
too-clever-by-half code-maker to the conclusions they wanted to draw in the 
first place) can be tricky to sustain. :-o

> Given that there exist such things as Mike Roe's generic word, and tables 
> of words which imply a limited number of prefixes and suffixes, and given 
> the unusually short word length, is it not possible that words in fact 
> represent relational ideas rather than literal words, or shorthand, (the 
> latter of which would not have so much redundancy, the former of which 
> would be far easier to decrypt) and the prefixes act as predicates, the 
> suffixes as arguments? This would also explain the apparent lack of 
> sentence structure.

IIRC, this has been discussed in regard to universal languages (ie, with 
words as serialised representations of positions within Ramist taxonomical 
trees) - but (IMO) the stats makes it a bit tricky to sustain. Can anyone 
give a guess how big the potential tree - ie, the dictionary - would have 
to be for this to make sense? For example, a six "letter" word would imply 
a choice at each one of its six letters - hence that one word implied a 
minimum of six choices in the dictionary... so the dictionary size could 
get out of control very quickly.

My guess is that it would rely on a dictionary of enormous size, probably 
twice as big as the VMS itself (just for the words, never mind any 
accompanying text), constructed like a gigantic tree - so such a hypothesis 
would require an even more extraordinary document than the VMS to exist in 
order for it to be correct.

Also, Ramus' original classifications were (IIRC) somewhat whimsical and 
arbitrary, far from the clean, logical arrangement one might imagine. Shaky 
foundations for a tall building, I'd say. :-o

> This idea of course would also work for the 'hoax' theory. It is possible 
> that were the VMS a hoax, the author may have invented a mechanical system 
> to create wording, avoiding randomness, which would make it obviously a hoax.

As any [compsci] fule know, real randomness is a hard nut to crack at the 
best of [computational] times - and 1450-1550 was probably not the best of 
times. Yes, the VMS could have been created mechanically - but (as I'm sure 
Gordon would point out) you should be able to detect the mechanism's 
signature in its products, not just from what it produced but from what it 
was unable to produce.

And why would a hoaxer avoid randomness? Roll four dice and a different 
code-table per paragraph and you'll get it over and done with - much easier 
than super-complex solutions. Anything else (especially over hundreds of 
pages) is overkill for a hoaxer, surely?

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


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