VMs: A theory about Wierdos

["GC" <glenclaston@xxxxxxxxxxxxx>]

As John has pointed out, the best approach is to
whittle away one concept or fact at a time.  This
concept concerns the handling of "wierdo"
characters, since they seem to appear in one form
or another on any given page.

I think we can all agree that no cryptographic
system can produce the Voynich in and of itself.
What hasn't been discussed to any great length are
systems that are fairly flexible and allow for
human manipulation.  A basic example of such a
system would be a homophonic substitution cipher,
that allows for multiple representations for a
plaintext character, solely at the whim of the
encipherer, but still within the confinement of a
table of replacements.

Many of you are also aware of other systems that
can produce a very high degree of repetition of
single characters.  If this can be done with
single characters, it can also be done with groups
of two or three if the enciphering table is based
on pairs or triplets.  We would of course agree
again that no system of this nature can produce
the Voynich in and of itself, without the
flexibility of human interaction.  But is there
evidence to support human interaction with the
system?  I believe so.

It's my contention that the gallows combination
characters, the G-form characters, and a few
others, along with the wierdos, are "interactive"
characters, and not individual characters in their
own right.  This also applies to the "in, iin,
iiin," set and all its variables, as well as the
various marks over the "table" character.  As for
standard characters with wierdo marks, consider
for a moment folio 49r, which has a series of
wierdo characters, not the least of which are a
pair of "4o"'s with tear drops drawn above them.
In my interpretation this would mean that the
characters that belong here are not "4o" at all,
but something else signalled by the mark.  The
"4o" are drawn only to maintain the appearance of
the text, and the mark gives reference to the true
pair of characters that should fill this spot.

If I am correct in my assumptions, the basic
Voynich alphabet would be around 23 characters,
with an additional 13 to 16 characters being
"instructional" versions of the basic characters.
This human interaction would allow for a much
greater degree of flexibility and provide for the
appearance of the Voynich script.

I will be working on this as time permits and
presenting some firmer evidence over the next few
months, but for now its enough that those
interested view this possibility in the light of
their own numbers.

GC