[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: VMs: Personal Guess

Hi GC,

I'd only like to point out once again that throwing VMS statistics out the
window without trying to explain them will only cost you time in the end.

My pair cipher hypothesis seeks to explain why stroke-level and glyph-level (and quite possibly word-level) statistics are likely to be unhelpful. ATM, I'm trying to regularise de-pairification so that I can convert from EVA to pairs algorithmically (so that I can run large-scale statistical tests on a de-pairified transcription) with some certainty, but nailing down the subtle differences between sections first would seem to be prudent. :-)

As to word-based statistics, you're dead wrong that there's no statistical
evidence in this regard.

I'd agree that every statistical test is "evidence" - but I'm just pointing out an alternate hypothesis which might explain why the statistical analysis done to date may simply have amounted to "knocking on the wrong door".

Are
you saying here that not only are the rest of us fooled and misdirected on
general VMS statistics, but that a study based on glyphs, due to the
author's habit of encoding his meaning in specific symbols is "next to
useless", the symbols we all recognize on the pages are now a misdirect?

Surely analysing the statistics of a fake mono-alphabetic cipher would be looking at noise?

Or
are you saying that the individual glyph is meaningless, that only the pairs
matter?  Either way, you're once again in a statistical limbo, but since VMS
statistics are misdirected, I guess I've a lot left to learn.

One of us is in a limbo, for sure - if you're sure it's not you, then so be it. :-)

Dismissing current
thinking on VMS statistics requires proving them wrong, not just saying they
don't matter.

I agree - but proper scientific study also demands transparency and openness, which I've tried to do, by posting where I've reached to along the way.

I'm certain I addressed this when I discussed the odd "sets of four" problem
I was having.  At the time, you were positive the VMS followed the Roman
Numeral track, so it's been awhile since I covered that "glyph" problem.  My
views haven't changed, but yours certainly have.  I'd cover this ground
again if it would help, but since glyph statistics are currently "next to
useless", I'll wait until they gain some respect. :-)

Tiltman's dain/daiin/daiiin Roman numeral hypothesis is by no means incompatible with my current prediction, which is that of a system composed of separate parts. So: in that respect, I'd prefer to say my views have "progressed" (as have all of ours, I'm sure). :-)

> After all, they practically invented modern cryptology.

That's some of the purest B.S. I've smelled since my last visit to Bush's
ranch! :-)  You've read some of my work on this subject, and you know that I
contend that the foundation of modern cryptography lies in the Qaballa, and
may date as far back as the 11th century in western terms of usage and
understanding.

Ermm... I did specifically say "cryptology", not "cryptography".

What the 15th century lacked over the
16th century is the open conflict necessary to bring about change, and
evolution is all about conflict and change.

Garrett Mattingly (in "Renaissance Diplomacy") argues at length that the 15th century (and specifically the 1453 Treaty of Lodi, following the fall of Constantinople to the Turks) forced the creation of cryptographic departments in Chanceries throughout Northern Italy. There was plenty of open conflict there, leading to the cipher evolution evidenced in the Milanese cipher ledgers (and elsewhere).

"the nature of steganography is to make the solution non-obvious."

Does the word "HIDDEN" ring any bells here?  Steganography, hidden writing,
is not designed to make the solution simply "non-obvious", rather to
*OBSCURE THE VERY FACT* that there is a message contained in the text.  200+
pages of blatantly obvious "hidden writing" makes no sense to me in terms of
"steganography".

If you're looking at the wrong alphabet for your statistics, then I'd say it is most definitely "hidden in plain sight" - so, a well-designed pair cipher is surely as much steganography as cipher, as would be a well-designed space-transposition cipher. We may have both of these here.

Your use here of the word "certain" is telling.  So in essence, we're all
wrong?

All I said was that I'm *personally* certain, I just have to find sufficient evidence to convince other people. The degree of sufficiency will vary from individual to individual. :-)

We're on solid
statistical ground, and you're in statistical limbo, but you'd have us
believe the reverse is true.

Well... I'd like to see the solution to which 91 years of statistical analysis have carried us, but apparently there's a big fat void there. But feel free to prove me wrong, I don't mind! :-)

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


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