Mon Jun 3 18:30:37 1996
From owner-tc-list Mon Jun 3 18:30:37 1996
Return-Path:
Received: by scholar.cc.emory.edu (SMI-8.6/SMI-SVR4)
id SAA23757; Mon, 3 Jun 1996 18:28:42 -0400
Date: Mon, 3 Jun 96 15:24:36 PDT
From: broman@Np.nosc.mil (Vincent Broman)
Message-Id: <9606032224.AA16905@Np.nosc.mil>
To: tc-list@scholar.cc.emory.edu
In-Reply-To: (hurtado@cc.UManitoba.CA)
Subject: Re: "Alexandrian" Text
Sender: owner-tc-list@scholar.cc.emory.edu
Precedence: bulk
Reply-To: tc-list@scholar.cc.emory.edu
content-length: 3100
-----BEGIN PGP SIGNED MESSAGE-----
hurtado@cc.UManitoba.Ca wrote:
> As to Colwell's suggested quantitative definition of a "text-type" (70%
> or better agreement in relevant variation-units), I've always taken his
> intention as somewhat pragmatic and ad hoc.
I agree. The problem I have with the "70%" is that we don't presently
have clear, universal agreement on how to arrive at the 70 numerator
and the 100 denominator.
The count of agreements will depend on how you divide up the units
of variation and on how you count subvariants. For example,
the Text und Textwert series sets up units of variation which can include
many tens of words and for which many tens of variant readings must be
listed, page after page. In such a unit of variation, two witnesses
might differ by some jot or tittle, or by some major rewriting,
and the TuT statistics merely count match=1, miss=0.
Any agreement statistics can be inflated or deflated by packing the
court, i.e. by being more or less inclusive in one's selection of
units of variation. Adding lots of minor variations where almost
all the witnesses agree anyway makes everyone look more similar.
In the bad old days the selected units of variation were just the
differences from the TR. Better analysts counted agreements
from among all the variants listed in Tischendorf. Colwell chose a roster
of witnesses which sample from all the known groups/families/types
and formed variation units from that representative universe,
possibly discarding singular readings (acc. to some definition).
Others have restricted themselves to variants listed in Nestle-Aland,
or even in the UBS editions. All these will come up with different
denominators in their percentages, which will then be compared against 70/100.
I like Hurtado's thinking in saying:
> This doen'st mean that W and P45
> aren't from a common "text-type" (in my view), but it does suggest that
> "text-type" has be be a bit flexible in what it means, or else W & P45 are
> (one or both) a bit more loosely members of whatever "text-type" one
> imagines them to represent than are Aleph & B.
It makes sense to me that one can formulate some consistent, repeatable
measure of affinity, and then report results in terms like:
"MSS X and Y are more closely related than W and Theta in Mark, but less
closely than 01 and B". Once this is established, then hypotheses
can be discussed as to how close they are in genealogical terms.
Vincent Broman Email: broman@nosc.mil = o
2224 33d St. Phone: +1 619 284 3775 = _ /- _
San Diego, CA 92104-5605 Starship: 32d42m22s N 117d14m13s W = (_)> (_)
___ PGP protected mail preferred. For public key finger broman@np.nosc.mil ___
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBMbNlyGCU4mTNq7IdAQHW/gP/ZM/gRI/8e0BZJuDvfWSfLdBKF7Yuqfxi
Nz5z9skxFyyyIF3OE9dGCwHHtICIW8zklwqPooFUJb3ExYj7xl+/uVbOOjeRd/Hg
yI603R91kLrFqCM3+7CpuOp/HxKQUVDNsliyqC7niA/ufvFtzMKSttce8A6MF7Jd
hkwC4EriKi8=
=fG3H
-----END PGP SIGNATURE-----
Back