Re: VMs: Image Source, Accuracy of Transcriptions

[Dennis <tsalagi@xxxxxxxx>]

Robert Antony Hicks wrote:
> 
> In doing this, I've found that the least 'matched' transliteration is Takeshi's,
> which disagrees with everybody else's interpretation as often as one
> character in ten, particularly gallows.  Whether this is due to Takeshi
> having better copies of the MS or worse eyesight I dare not speculate.

	I have no idea of why this should be; see below.

Rene Zandbergen wrote:
> 
> The accuracy could only be established by
> going back to a good-quality copy, e.g. Friedman's
> originals.

	I didn't think Friedman had good-quality originals! I
now remember that there are very good-quality B&W
copies at Yale in the material along with the VMs
itself.  Are these the ones in question?  

	In the archives I found Frances Wilbur's reminiscences
of the FSG's transcriptions efforts.  I wonder why I
heard such a poor estimate as 5% of the FSG versions
quality?

> By its nature, the majority transcrtiption should
> be the most accurate one.

	If they all agree, yes!  EVA captures rare characters
like the picnic table and also complex ligatures and
'weirdoes' which FSG and Currier don't capture, so the
Takeshi's version is better there.  It sounds like
Friedman had the best quality copy, then Takeshi, then
D'Imperio.  After this discussion, it's not clear to me
who took the most care!  I think one can say:

1)  If all three agree, that's the best.
2)  Takeshi is the only one who captured the
distinctions mentioned above (but did he?  I haven't
looked).
3)  If all three disagree, it's completely
indeterminate.

	Beyond that, I can't even speculate.  Yet more light,
anyone?  

> The real problem is that in order to assess the
> accuracy, one really has to know which features
> of the script are relevant and which ones aren't.
> For example, GC is of the opinion that the different
> curls in the various shapes of Eva-sh or
> Currier-Z are relevant, while none of the existing
> files makes any distinction. Similar questions
> could arise over the shape of the tail of
> Eva-y / Currier-9 or the various intermediate
> forms between Eva-r and Eva-s (Currier-R and
> Currier-2). If these details are relevant, then
> none of the transcriptions are of top quality,
> perhaps not even of sufficient quality.
> 
> We can only hope that they are not.

	Indeed, the glyph assumptions!  After having talked to
Barbara Barrett, I'm wondering myself.  We seemed to
have reached a dead end; could it be because we aren't
in fact capturing relevant distinctions?

Dennis

--------------------------------------------------------
Subject:  Voynich reminiscences of Frances Wilbur
   Date:  Mon, 6 Jan 1997 08:25:46 -0500
   From:  "Jim Reeds" <reeds@xxxxxxxxxxxxxxxx>
     To:  voynich@xxxxxxxx

On Friday 27 Dec 1996 my wife and I had a several hour
coffee chat
with Mrs Frances Wilbur and her husband Dr William
Wilbur.

<snip>

She remembers only some things about the FSG's
activities, possibly 
because most of the FSG's members were from the "C"
section (cryptanalysis
of enemy systems) whom she did not know.  She was
present at the initial
meeting of 22 people when the alphabet symbols were
chosen.  The FSG had
no special name for itself.

She remembers working with her partner Salome Betts
transcribing VMS
text from the large-size photostats now in the Marshall
Library.
Transcribing was done as follows.  After hours she and
Salome met at
Frances's apartment, waiting for a time when her
apartment mates were
absent.  (The FSG members had been told not to talk
with outsiders
about the Voynich project.)  Then, at a table, she and
Salome would
transcribe.  One would hold her finger to a Voynich
letter in the
photostat.  The two agreed on the transcribed value and
the other
wrote the transcribed value onto the sheet of quadrille
paper.  The
work was incredibly slow, but not unpleasant.  ("We
thought it was a
ball."  "You had to have a sense of humor or you'd go
bonkers.")

She could not account for the use of alternate
transcription alphabets
(the "alphabetical" and the "mnemomic", in the terms of
my paper).
She could not recall how overall checking was to be
accomplished, nor
whether other transcribing teams were assigned the same
pages as a
check.  (It's possible that two transcribers at once
was thought good
enough.)  She did not know how the transcribing tasks
were divvied up,
but thought that "Dusty" (Capt. Marc) Rhoads would have
been in
charge of that.  She must have done a lot of
transcribing work,
because her name appears not only on the "1613"
transcription that I
mentioned in my paper, but also because her (first)
married name, Mrs
Wilford, appears on the covers of another bunch of
printout material
in the Marshall Library.
--------------------------------
[This E-mail scanned for viruses]

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