Re: VMs: Strange or not?

[Eric <mynumberis2000@xxxxxxxxx>]

--- Koontz John E <John.Koontz@xxxxxxxxxxxx> wrote:

> Just to reassure you both, I had realized it wasn't
> Knox who was
> suggesting the forms were allo... alloglyphs?  My

I'm probably using a different term from a different
discipline (cryptography) for possibly the same thing
- homophone. In a homophonetic substitution, there is
more than one possible substitution for the same
letter in the plain text. Unlike a polyalphabetic
substitution, where generally every plain text letter
has the same number of alternate substitutions and the
substitutions are used in a fixed order (possibly
based on a rotating keyword), in a homophonetic
substitution, each plain text character has a varying
number of substitution characters and the order in
which they are used is random (up to the author's
whim).


> I think overdifferentiation is safer as long as
> there's no secure way to
> determine if forms like shol and chol actually
> contrast.  Lowercase bit :
> pit, bet : pet, rib : rip, brick, prick, etc., might
> be tempting to
> conflate if we didn't have glosses.  There are
> distributional differences
> for b and p, too, e.g., spit, but *sbit.

Right, so one has to be careful. I'd want to see a
series of phrases and not just single words. For
instance, seeing something like this happen often
(randomly made up series of letters representing a
common plain text English phrase of "and the"):

basil ample : basit ampte

Would lead me to think maybe l and t both represented
the same underlying letter.

Anyways, I haven't done any of this statistical work
to really try this out, it was just a thought that
flitted through my head when seeing some of Knox's and
Marke's work.

Guessing investigations of this sort have been
attempted before... searching Google for "voynich
homographic" and "voynich homophonetic" turns up only
an Andras Kornai paper which doesn't seem to directly
deal with cryptography....?

Overdifferentiation can hurt. The base text needs to
differentiate as much as possible. But it is important
to start removing these differentiations when we have
a belief/theory/guess that they are non-existant. Word
lengths, letter frequencies, consonant/vowel
algorithms, guesses at word parses and grammer - they
all depend on a correct representation of the actual
differences between letters and words in the text. 

My "best practice" outlook is to start with the most
differentiated file and then document each aggregation
made.

I'm hoping to have a second try at my aggregation
program next week (my first iteration months and
months ago was very clumsy). If something interesting
pops out, I'll pass it along. And if I don't post
anything, well, then you know what happened ;-)


> The difference between EVA c and s is a lot more
> graphically deliberate
> than that between t and k, as I think everyone would
> agree, but the
> general character of the VMs script is such that it
> seems likely that
> subtleties count.

Maybe they count... What I mean is, I fully agree they
are graphically deliberate differences - not only
between c and s but also t and k (looking at the
actual scanned images at instances of t and k, it is
clear to me that the author meant each one as written
and that they are not just sloppy versions of each
other due to variances in writing). However, that
doesn't require they mean different things if we are
going for a cryptographic explanation - that's how a
homophonetic cipher works. Looking at some of the
Trachedino ciphers, very similar looking characters
were used to represent the same underlying character.


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



		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - Helps protect you from nasty viruses. 
http://promotions.yahoo.com/new_mail
______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list