[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
VMs: GC's request on Currier
--- GC <gc-@xxxxxxxxxxxxx> wrote:
> Sorry to wake you, :-)
Having one's mobile go off when E-mail arrives
is one of these newfangled things I can live without
(certainly for the VMs mailing list), so there
is nothing to apologise for :-)
> Your Currier "hands" listing demonstrates Currier's
> identifications, but there are large gaps in the
> information. Correct me if I'm wrong, but
> didn't you do extensive work in this area, and if
> so, can you offer a brief chart, similar to the
> Currier chart, that summarizes this effort?
> [ ...]
> Can we use the labels we discussed last year that
> do not involve the words "hand" or "language" to
> describe statistical groupings? It's grown
> beyond Currier, and as much as I think of him,
> some of the misleading terminology should be
This is a straightforward question, but a
straightforward answer is not possible. It is,
however, possible to be at least a bit systematic
in answering this from my viewpoint:
- Currier identified 'language' and 'hand'. The
first distinction is based on clearly defined
and verifiable criteria, but the term is a bit
unfortunate. The second distinction has a good
name (it does refer to the appearance of the
script) but his classification could be challenged.
- Currier did not analyse the entire manuscript,
but presumably only those pages for which his
classification is now known.
- We can do much better now (and more easily so
on the 'language' criterium) so I do see his
results as something of historic value rather
- As a result of my own analysis, I do not support
the dual-'language' hypothesis of Currier, but I
do understand that his conclusion was correct for
the subset of the data he had available (i.e. not
the entire MS had been transcribed).
- I could attempt to classify the pages of the MS,
based on my analysis, but I would not be able
to state the classification criteria as clearly
as Currier did.
As is probably known, I did the analysis on the
basis of digraph statistics (not in Eva, but in
something closer to Currier). Stolfi has done
something similar, based on word statistics.
My main conclusion was that the statistics show
'overlapping clusters' where the text statistics
of pages with the same type of illustrations
tend to be rather similar (with the notorious
exception of the two types of herbal pages).
I can and will do it, but again, it won't
be as simple as 'A' vs. 'B'. I have never looked
at 'hands', but I am of the opinion that it
is quite a promising avenue for someone to see
if there is some kind of progression in the hand-
writing style in the MS (not necessarily in the
order in which the pages are now).
Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying: