Re: Evolution - was VMs: Inks and retouching

[Koontz John E <John.Koontz@xxxxxxxxxxxx>]

On Sun, 1 Aug 2004, Gabriel Landini wrote:
> > Might I also ask, as a latecomer, what the b is, structurally?
>
> Same as above, this is also a "new" character. It is like an all curved <n>.
> It is almost always word final, and preceeded by <e>, <ee> or <eee>. There
> are a few cases of <b> in my reading. It may have been transcribed previously
> by others as <n>.

Thanks for the eludication!  It's difficult to search for "b" in an
archive ...  Although using Grove's analysis b looks more like a variant
of o - pace Grove! - it would make sense for it to be treated as n, too.

> <f4v.8> oleeb.chor...
> <f5r.3> qoaiin.[o|a]tan,chy.daiin.oteeeb...
> <f7v.4> choteeeb...
> <f14r.12> oeeeb...
> <f20r.5> pchocthy.chokoaiin.(cp)y.cheeeb...
> <f23v.11> y,okaiin.doroiin.olols.oiin.-ol.cheeb...
> <f29v.9> sho.chokor.chor.chy.ydaiin.cho.-ykeeb.-...
> <f30r.11> qotchor.cheor.chey.cheor.chey.soee[b|n]...
> <f35v.18> daiin.qokeeeb...
> <f76v.37> qoeedy.lchedy.cheeb...
> <f88r.5> yokeody.cheom.qoor.cheeb.y
> <f102v2.9> oldeey.ckhy.sheeody.eeos.she[b|sh]e.
> <f106v.30> otaiin.okochey.qody.oeebysarx...

> > In essence Grove proposes that EVA approach misanalyzes the glyph level of
> > the Voynichese script, but not looking hard enough at the way in which the
> > graphs are patterned.  It stops short, fooled by the lack of connections
> > in some cases.
>
> I am not sure it "misanalyzes" because the purpose of it was not the
> analysis, but to help transcription. The "final" alphabet is still open
> to debate. This has been covered too many times in relation to other
> character groups:  whether <sh> is <c'h>, <qo> should be a single
> character, etc. so I will not go into that again -- that is what mail
> archives are for.  :-)

This is all true, of course, and I didn't by any means mean to imply that
any misleading qualities of the transcription were deliberate or even
illconsidered.  It's just that some of the patterns Grove adduces are
somewhat obscured in EVA and its predecessors.  I was, in effect, warning
myself against a tendency to see the glyph analysis of the EVA as atomic
and final.  There's a certain charm to "fachys ykal ar ataiin shol shory
chtres ..." that tends to overwhelm one!

> But, yes, it is possible that <n> may indicate that the previous (<e,i>)
> characters are part of the same group.

And, I would imaigne, a particular kind of group.

> This opens further questions: what are <i> and <n> on their own? (i.e.
> no <n> after <i> and no <i> before <n>?). Are these the same character ?

Well, I would hypothesize that one is the analog of the basic consonantant
elements of a consonantal script (somewhat artifically distinguished into
abugida and abjad types) and the other is the analog of the vowel
diacritics employed in such systems.  In some sense, neither element would
stand on its own.  Currently it looks more likely to me that the n-like
element (or the various hooks and loops attached to the last c/e or i) is
the vowel diacritic and that the c/e and i lists or bundles represent the
consonants.  (And then there are those awkward other things.)

I rather wonder if the c/e and i bundles might not amount to references
into some sort of orderly table of consonants and/or clusters of
consonants, but that would be asking a lot of Fortuna.

The feasibility of this will be clearer to me when I have had a chance to
try to ennumerate the possible c/e and i bundles underlying the EVA
transcriptions.  If that enumeration doesn't lead to something apparently
useless, then the question is whether I can think of a computationally
feasible way to test with substitutions.  I hope that the structure of the
c/e/i bundles will suggest something, or the contexts in which various
retranscribed forms occur will do this.  (There's no end to my optimism,
is there?)

> This does not solve the parsing of <iiiin, iiin, iin, in> either:
> iiiin = ii+n or i+m?, etc.

I'd say i*n and i*m (using * for '0 or more of the preceding') are two
distinct sets of things, and that EVA n and m per se each include already
one of the i-graphs, with an attached hook.  In the case of n, that hook
is attached to the bottom extreme of the i (or e/c) and curves out
and up to the right and then back over to the left.  Attached to an i it
produces the EVA n.  Attached to a c/e it produces the EVA b and - I would
suppose - o, though this is an area where I am somewhat uneasy.

In the case of m, the hook is attached to the top extreme of the base i
(or e/c) and curves up to the right and then crosses back over itself,
descending parallel to the base character and finall curving leftward
under it.  Attached to an i it produces the EVA m.  Attached to a c/e it
produces a g.  There is some possibility that m and g are more flamboyant
variants of j and d.  For now I'm assuming not.

As far as the bundles, we see things like c(1:4) (1 to 4 c's), i(1:4),
ci(1:4), and so on.  If I'm right, these are consonants.  (If I'm wrong,
my batting average isn't any worse than anyone else's, though my personal
mortification will be real enough.)

In these terms fachys ykal ar ataiin shol shory is something like follows,
in which each minor group (within a word) is a syllable or unsyllabified
stop.

f  ach  y  s  . y  k  al   . ar   .  a  t  aiin   . sh  o  l  . sh  or  y
1j cich cl cr . cl 1m ciil . ciir .  ci 4m ciiiin . crh cn il . crh cin cl

If, just for jollies, I assign aeiou to the rnjlm hooks, the syllables are
something like the following, using ??? for aspects of the system I don't
have a hypothesis for.

Ci???CoCa CoCuCu Ca ???CuCe ???CeCo ???CeCo

Not very edifying, I admit.  I hope I've done the substitutions right!

> For example, in Spanish for "penguin", to make the "u" sound after a "g"
> in "gui, gue" we use a "dieresis" (not sure how this is called in
> English) on the "u": pingüino.  For somebody who does not know Spanish
> (and not guessing what a pingüino is!), it is impossible in a hand
> written text to know whether this is "iii", "üi"  or "iü". The Spanish
> reader has no problems because the word is read in a context (you also
> know that dieresis on the u, if preceded by g, is followed by i or e, so
> "iü" is not an option unless it is a foreign word.)
>
> What I mean is that this kind of problem may not be easy to resolve since we
> do not know the underlying language (if any!).

If situations of this nature occur, they will present problems, of course!

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