Re: VMs: ... speculating with "dairol"

[Dennis <tsalagi@xxxxxxxx>]

Koontz John E wrote:
> I believe that the Voynich ms glyphs rendered as vowels in the EVA
> transcription consistently behave as vowels in statistical analyses of
> vowel-like behavior.   

	Gabriel did a chart that shows CV identification by the 
Sukhotin algorithm for different transcription systems.  I 
can't find anything about it; it was on his site, which is 
down at the moment.  I think you are basically right that 
EVA vowels are Sukhotin vowels.

> I do not know to what extent similarity of shape and
> patterned behavior determined the equation of Latin characters to Voynich
> glyphs, but it seems to me that both were to some extent deliberately
> employed.  Apart from this, the various tables of transcription systems
> available on the Web show that there has been some degree of mutually
> influencing evolution in the development of such systems.

	Yes.  There's no need to speculate, though.  Gabriel 
invented EVA, ask him!

	All the systems make the transcription symbols resemble the 
glyphs for mnemonic value.  I'm not sure whether anyone 
considered patterned behavior.

> In spite of this it seems very important to remember that EVA is not
> Voynichese.  The encoding is very convenient, but it may be misleading in
> various ways.  For one thing, it implies an analysis of the Voynich
> writing into character entities that is very likely to be incorrect.  For
> another, it does suggest vowel and consonant readings.  In fact, the EVA
> transcription is perhaps a bit too seductively "pronounceable" to be quite
> safe!

	-iin is obviously, in most cases, a single glyph.  To 
consider it a double vowel followed by a terminal consonant 
would be a big mistake, one I'm seeing too many people make.

> Anyway, even if you accept the resolution into characters, even if you
> accept the consonant vs. vowel distinction implied and the clever way in
> which the peculiar bench and gallows patterns are encoded with (c,s) + C +
> h sequences, it seems unlikely that all the matchings of particular
> consonants and vowels are correct.

	More generally, the VMs most likely contains digraphs; 
several glyphs may represent a single underlying phoneme. 
Or some cryptographic equivalent thereof.  Surely we're 
missing this in EVA and all the others.

	But none of the transcription systems were ever intended to 
be taken at face value!  They were all simply intended to be 
the starting point for further analysis.

> > My questions are: Can the fact that EVA is somewhat pronounceable tell
> > us anything helpful at all about Voynichese? Can the failure to resolve
> > the glyphs into truly pronounceable words tell us anything helpful about
> > Voynichese? Can the attempt be of any benefit?
>
> To me it suggests that the encoding - no matter what the details are -
> preserves the syllabicity (CV alternations) of the underlying language and
> represents consonants and vowels in different ways.  That is, an
> underlying form like cat (or whatever) is not encoded so a constituent
> element in the encoding can sometimes represent c, sometimes a, sometimes
> t.  It seems likely that the representation of a given sound is always the
> same, or at least that consonants are usually represented in one way and
> vowels in another.

	See the remark above about digraphs.  I think the VMs also 
involves some type of cipher, that it isn't a 
straightforward representation of a natural language.

	Jacques pointed out to me that the Sukhotin algorithm 
doesn't distinguish consonants from vowels, but rather sets 
of symbols that do and do not contact each other.   I assume 
that other vowel identification algorithms work the same 
way.  They are a standard tool for breaking simple 
substitution ciphers.  I'm not very sure about what these 
algorithms say about unknown languages/scripts.  Remember 
too that scripts may contain logographs or symbols for 
several phonemes, as Egyptian hieroglyphic did.

	I don't see why consonants should be represented on one way 
and vowels in another; that isn't true of the Latin alphabet 
so far as I can see.   The Latin vowel symbols ultimately 
came from proto-Semitic consonant symbols.  /A/ came from 
the symbol for the glottal stop, an Egyptian symbol that the 
Sinaitic Semites thought looked like an ox head.

> I'm not sure exactly where Tolkien got the basic ideas for his script, but
> he was fascinated with languages and familiar with the classical
> Indo-European ones, and his scheme bears a certain generic resemblance to
> the approach used in Devanagari.  It also resembles to some extent the
> medieval and later approach to abbreviating words or parts of words to
> consonantal frameworks with attached diacritics for omitted vowels or
> inflectional endings.

	We haven't given much thought to possible diacritics in the 
VMs, except for the hooks over some of the benches.  You 
make an interesting point about Medieval abbreviation 
systems; Capelli has been our resource for these, and maybe 
Bischoff.  EVA /y/ looks like the abbreviation for the 
common masculine singular -us .  But /y/ looks much more 
like 9 and is not a diacritic.

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