Re: VMs: Viola tricolor

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi William,

At 23:27 26/05/2004 +0100, William Edmondson wrote:
> That indeed is an interesting question.  But not actually relevant to the 
> document being a hoax.  Despite the difficult of proving the VMS is a hoax 
> one could probably reach a 'balance of probabilities' conclusion on that 
> without knowing for sure who was hoaxed.

IMO, there are two pragmatic ways to "prove" that the VMs is a hoax: (1) 
unearth provenance information that implicates it directly as part of a 
hoaxing scheme, and (2) reconstruct the method by which it was generated.

Though some people assert that the suggested connection with a known 
counterfeiter (Edward Kelley) is sufficient for (1), the actual evidence 
seems to fall well short. As for (2), some list-members belittle Gordon 
Rugg's efforts, but AFAIK his work is the first relatively systematic 
attempt to see if the VMs was hoaxed using a specific method.

In the end, though, I think the idea of VMs-as-hoax sits uncomfortably with 
the observed multiple levels of structure in the text - letter-adjacency, 
word-structure, line-structure, paragraph structure, page structure, Neal 
keys, etc. Unless they are all artefacts of the way by which the document 
was constructed, why would they be present in a hoax? Randomness is very 
hard to sustain: unless you can show *how* the VMs was hoaxed (by 
reproducing the kinds of textual behaviour we observe), the balance of 
probabilities is that it is indeed structured to contain meaningful 
information. All credit to Gordon for trying to do exactly this: though I 
personally suspect he is trying to re-hoax too small a subset of 
Voynichese. :-(

Just so you don't think I'm kicking hoaxes alone, these same issues cause 
problems for the idea of VMs-as-language: why would (for example) a natural 
language contain Neal keys, no obvious numbering system, and have 
paragraphs (nearly) always starting with gallows characters (never mind 
split ol pairs, etc)?

By way of contrast, the key problem with the idea of VMs-as-crypto is that, 
for it to be true, we have to accept that one person 500 years ago was a 
crypto genius - all the statistical artefacts we observe are consistent 
with a signal's being hidden inside the text, using tricky ciphering 
mechanism(s) we're not ~quite~ sneaky enough to break. We can accept a 
Brunelleschi through his works: so why can we not accept the VMs' cipher 
designer too?

> I take it that no-one knows of a comprehensive mutilingual attempt to 
> examine labels in the context of plausible interpretations of labelled 
> images?  It really would seem the best basis for independent stabs at 
> meanings to be attached to labels.

I did once try this, but every plant I looked at had both a large number of 
translated (language) names, and a huge number of local 
(dialect/regional/district/town/individual) names. Unless you have a very 
good idea of both language AND locality in the VMs, collating all this 
could well be a life-time's work. So: while it's "possible", the practical 
difficulties involved are surprisingly high. :-(

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


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