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

RE: Glen's Rant



John, thanks for being so kind.  Pinyin is the
name of a pet pig on Disney, right? :-)

The in, iin and iiin problem is only one of the
few I'm trying to get to the bottom of.  Even with
36 characters I'm still getting counts of around
23 or 24, and that's most interesting to me.
Setting up a chart in PDF is the only way I can
demonstrate the grouping of these stranger
characters, since I detect a pattern in their use.

I have a sense of the meaning of ligatures, tails
on o's, curlies moved to the right or left of the
picnic table, etc.  I have a field in the database
that I mark for these variations, and there just
aren't that many percentage wise.  When and where
they show up is another matter.

If your alphabet for a particular language
contains 24 characters, you can reduce your
character set to 12 and still viably communicate
information, especially in cipher, so size of the
character set is not really as important an issue
as what an actual character is.  If we look at
some of the uses for the Masonic Pig-Pen cipher
with its 9 angles, we see several times that it
was used without defining marks, meaning
information was transmitted with only 9
characters.  A bit difficult for the end-user to
decipher, but still intelligible with some effort.
If the VMS has 100 characters that's fine with me,
as long as I can determine what the unit of
character should be.

As an example, the characters that look like half
a picnic table on either side of a gallows
combined only occur 8 times in the first 45,000
characters.  That number in itself gives you the
odds of these two actually being unique
characters.  An extremely low confidence level
that these are characters, and conversely an
extremely high confidence level that they are
variants of standard characters, possibly a "c".
Do I assume it's a "c" and move on?  No, I have to
look at other characters on the pages where these
occur and determine if there is some pattern to
their usage before I can make a determination as
to meaning.  It could be a truncated gallows
combination, and this can only be determined by
plotting where these show up and in what context.
Judgment isn't perfect, but it can be augmented
with the tools available to make it the best
logical choice.

How often does dain, daiin, daiiin appear, and
when you blow the character up to one-inch size on
your screen, are there connecting strokes between
the i's that have faded over time?  Many of the
curls above the picnic tables are so faint as to
only be barely perceptible on the color images,
but they're there.  They're certainly not the only
strokes that have faded over 500 years.  In my
estimation the best way to proceed is to examine
each example carefully in the best light, count
the number of times they appear to vary, and form
an educated guess based on the numbers.

In the final analysis it is paramount that the
best judgment be made so the most accurate
information is available to work from.

GC





-----Original Message-----
From: John Grove [mailto:John@xxxxxxxxxxxx]
Sent: Sunday, February 24, 2002 9:03 AM
To: Voynich@Rand. Org
Subject: Re: Glen's Rant


Wow!

    Funny, intuitive, condescending, and
motivational! Interesting
combination.
Where does one begin to respond?

    Well- I guess: The facts of VMS Characters
(not the characters studying
the VMS)...
> The G/picnic table combination has two styles,
one
> with a full looped G and one with no leading
loop.
> My sense is that these are the same character.

    There are slight variations that may be same
character, but can we
safely say they are - because those variations
show up a lot.

> When an N occurs at the end of a word, it has a
> tail, but when one occurs in the middle of a
word
> without a tail, is it a different character or
> still an N?

    What's an N without a tail look like? Are you
saying two (god forbid
EVA) i's make an N - even if there is no ligature
finishing the character
off?
so (you gotta love it) daiiiin would be 8ann?

>
> I think it's interesting that new characters are
> only introduced on the "nocturnal time-piece".
I
> read an earlier posting where someone thought
> these might be actually related to zodiac signs,
> (I'd say more probably certain stars in a zodiac
> sign).  I don't remember who sent the post, but
I
> think it was a pretty good call, and in line
with
> my own observations as to the meaning of this
> diagram.

    First, it's really coincidental this
'nocturnal time-piece' discussion
is taking place now - my son just did his Science
fair project on building
one to tell time with Polaris, Dubhe, and Merak!

    I'm not sure if someone correlated this page
with zodiac signs before -
I don't recall. However, f68r3 may mark the four
seasons with stars from the
zodiac signs...

>
> As I said earlier, it is very necessary to
> evaluate these patterns and determine the actual
> unit of character.

    Agreed - problem though... what happens when
you have
characters like a perfectly formed 'o' that have a
ligature on top
of them? Slight variations of this sort happen
throughout the document -
including the 'n' that is fixed snug up against an
'a' to look like an @.
For 'n' alone - we have @, an, ain, aiin, aiiin,
and aiiiin(I think).

    If you make in = N, and iin=M, then what are
the @ and aiiiin types?
    There are a number of characters that finish
with a ligature that they
shouldn't - by strict character definition. 'chy'
where the 'y' isn't
separated from the 'ch'...  This is why I favour
some kind of ligature
'vowel or doublet' marking system - but
unfortunately that reduces the
character far below acceptable...

>
> I think there's a way of statistically
determining
> what a character unit is, what is a valid unique
> as opposed to a simple variant.  (I've been
> striving to eliminate variants for some time,
but
> hindered by the poor quality of the copyflo.
> Fortunately this last year or so has made
> available approximately half of the VMS in image
> format, which allows much more reliable
> verification.)

> Anyhow, someone posted an
> example of a message board that accepts font
> values.
http://www.discusware.com/discus/index.php.  One
> of my favorite simpler "skins" is viewable at
>
http://www.zodiackiller.com/messageboard/messages/
> board-topics.html.

    I posted a sample form that allows users to
input using a font
installed on their system (and the server). I'll
check out the message
boards
you reference above, but haven't heard whether
most users would want to
migrate to a message board environment over the
email. In fact, this is the
only message that alludes to the use of the
open-forum that I've seen since
posting the sample form.

> This is a very good middle ground for those who
> are frustrated with the UNIX-Vision of
monospaced,
> mono-typed text adhered to by those who go
> ballistic every time you forget and leave your
> e-mail set to HTML.
> What good is the greatest aid to
> communication in human history if it's stuck in
> 1980 Wang-Word-Processor Mode?)

    Nice (and Valid) Rant... Similar to one I've
used in my work environment
that includes people who are using very obscure
web browsers (or very old).

> that the "greatest" cryptologist in history
tried
> and failed, as if discovering the key to an
> astrological cipher was that great and
> insurmountable a mental feat.  Our imaginations
> make this far more baffling that it should be.

    Our imaginations are our biases - yes. As for
the motivational challenge
that the geniuses haven't solved it, well -
sometimes genius is simply
seeing the simplicity that everyone else has
excluded 'because if it was
that simple - it would have already been
solved/inventend/discovered...'

>
> Meanwhile, performing entropy calculations on
pen
> strokes certainly keeps the mind occupied, but
> unless the entire 10,000 characters of the
Chinese
> language can be reduced to 23 or 24 distinct
> character forms, this popular activity is of
> little useful interest.

    What is pinyin?

> mentioning "dain daiin".  Both patterns are
three
> characters, not four or five, (8an and 8am).
Why
> make life so difficult?

    Because there are too many forms of 'n' to
simply say
in = n and iin=m (per above).

>
> The standard argument is that we have no idea
what
> the language is or what constitutes a character.
> On a scale of dung to crap I give that argument
a
> BS+.  You have hundreds of characters throughout
> the manuscript standing all by themselves,
> screaming "look at me, I'm a unit"!

    Nowhere do I see an 'ii' by itself to say
scream "I'm a unit - without a
tail"

>  I focus on the "systematic" as a
> good thing, meaning there is some method or
system
> behind this to be discovered.  If you choose to
> focus on the "unintelligible" as a bad thing,
the
> "system" has achieved it purpose and you're off
> tallying pen strokes.

    Now, now... Systematic is a good thing - but
why say tallying
pen strokes isn't systematic? You're doing it
yourself when you said that
you're trying to reduce your character set from 36
to 23 - excluding
'variations' (what are the variations but
additional penstrokes that don't
fit your 'basic' character).

>
> I'm not here to argue with those of different
> minds this time around, as I believe we're both
> entrenched enough that this is pointless.

    Yes, we are human and subject to our own
biases, even when we feel we're
open minded. The only way to bend our individual
mindset - is to nudge us
slowly by small factual statements that we can
concede the point.
The more points you score - the more our biases
are realigned with yours...

> This is only the means by which
> Nervana is obtained.  I strive only for the
Pizza.

    Nice finishing touch. Thank Jorge for bringing
in the award!

John Grove.