Mon Apr 29 16:38:03 1996

From majordom  Mon Apr 29 16:38:03 1996
Return-Path: 
Received: by scholar.cc.emory.edu (5.0/SMI-SVR4)
	id AA07518; Mon, 29 Apr 1996 16:38:03 +0500
Date: Mon, 29 Apr 1996 16:35:00 -0400 (EDT)
From: Maurice Robinson 
To: tc-list@scholar.cc.emory.edu
Subject: Re: Byzantine Text
In-Reply-To: <9604292009.AA19840@Np.nosc.mil>
Message-Id: 
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Content-Length: 2153
Sender: owner-tc-list@scholar.cc.emory.edu
Precedence: bulk
Reply-To: tc-list@scholar.cc.emory.edu



On Mon, 29 Apr 1996, Vincent Broman wrote:

> For any kind of evaluation involving Byzantine witnesses,
> the selection of readings in UBS3 or even NA26 is highly unrepresentative,
> and will make your counts of agreements nearly meaningless.
> For instance, in this case lots of units of variation where Westerns
> and Alexandrians agree against Byzantines will not be covered by
> any entry in the apparatus, because the text is felt to be firm.
> This is a standard pitfall in using a selective apparatus.
> You need to work from the IGNTP volumes, Tischendorf, or Von Soden.

I might interject that since both Waltz and I did work from the UBS 
apparatuses, that in my Th.M. thesis a strong justification was made for 
the legitimacy of using those variant units (at least for the Gospels and 
Acts) as sufficiently representative statistical samples, especially 
samples which would not be as likely to be skewed as those minority 
readings which dominate the N27 apparatus, where virtually everything 
else reads with a common text half the time.  

Metzger concurred with both my method and results in a private
communication made back in 1975, and concluded that the results were quite
in conformity to those which might have been obtained by a more precise
examination of two or three selected chapters via the Claremont Profile
Method or Colwell's Quantitative Analysis (followed by J.G.Griffith in 
his "Numerical Taxonomy" studies, e.g. JTS 1969). 

Thus, I do not think that either myself or Waltz' results will be 
significantly different, taken and understood as a representative sample, 
from those which would occur were a complete analysis performed using a 
full continuous text model as Broman urges.  There is a standard 
deviation, and a certain margin of error, but once this has been allowed, 
the results should be meaningful.

_________________________________________________________________________
Maurice A. Robinson, Ph.D.           Assoc. Prof./Greek and New Testament
Southeastern Baptist Theological Seminary     Wake Forest, North Carolina
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Back