Re: VMs: Follow-up

["surfing t" <surf17@xxxxxxxxx>]

>IMHO the strongest argument against a hoax is: The closer you look at the text,
>the more rules become apparent which govern the composition -- like "Paragraph
>initials are gallows", "serveral glyphs are word-inital or word-terminal only", etc.

Yes I agree. I'm also impressed that the text apparently satisfies some statistical laws of languages. However, as an amateur I have no way to know the significance of this. It would be helpful if someone knew the probability that random gibberish satisfy the various laws equally well. Even if random text is unlikely to satisfy the laws, it doesn't rule out the possibility that the VM was created using some algorithm the output of which, for some subtle reason (unknown to the supposed hoaxer), satisfies many of these checks. Of course the more checks that we find the VM's satisfies the less likely this becmes. In that case, even if the VM is a hoax, recovering the algorithm would in itself be interesting (it's probably possible to construct such an algorithm with today's knowledge, but this would have to be a simple algorithm that didn't purposely try to satisfy the laws). As far as I understand, the Cardan grille produces text that satisfies some of the properties, but this!
 in
 itself doesn't explain why Zipf's law is satisfied as far as I can see. Are there any other laws that Cardans grille doesn't explain?
If we take the Zip-law as proof the VM isn't just gibberish, one would either have to accept the VM as being unciphered or that it was ciphered in some strange way that preserved the word frequencies - in that case, at least we should be able to rely on the frequency of words to be accurate but the internal structure of the individual words may have been scrambled. But apparently there is some internal structure that does seem like a natural language, so the cipher would have had to preserve this as well meaning it must be very simple! If that's the case, the original language must also have been unknown to us (otherwise the cryptographers would have found out) and in that case the cipher is in a sense invisible to us, because the resulting language is not more difficult to crack than the original unknown language.

>Besides, there even are
>features (like the high word repetition rate) which make the VM look
>"unnatural", and which a hoax author would probably not have used for this very
>reason.

I don't totally agree with the above argument, because if it is true, it invalidates itself: The trickster could equally well have carried out the above reasoning and would thus have decided to include the repetitions on purpose, knowing that the finder would with high probability take the repetitions as a sign of the document being genuine. 
Of course this line of reasoning is problematic too: If the VM had contained 1000 repetitions of the same word, everyone would rightly be skeptical. For the "repetitions trick" to work the hoaxer has to make the document look otherwise plausible', and he has to limit the number of fake repetitions in order to maintain a balance between what seems a bit out of the ordinary (thus leading the reader to be suspicious and then eventually carry out your reasoning) but not too far-out - so it's kind of a gamble. If I were to create a fake book I'm not sure if I would risk it because even if it works out the effect is quite small. In case of the VM, I think the repetitions in general have done more harm than good in making people believe in it. So maybe the hoaxer failed in his reasoning or maybe it isn't a hoax after all. I'm confident however, that if it is a hoax, the hoaxer himself have been fully aware that he was doing the repetitions and the amount of suspicion it would raise!
 an
d he must have done so for a purpose. 

In short, I don't think the repetitions in this case, says anything about the document being a hoax or not. 

Thanks for your reply :) Have a nice day.


____________________________________________________________
Find what you are looking for with the Lycos Yellow Pages
http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10
______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list