Mon Jun 17 09:20:53 1996

From owner-tc-list  Mon Jun 17 09:20:53 1996
Return-Path: 
Received: by scholar.cc.emory.edu (SMI-8.6/SMI-SVR4)
	id JAA23551; Mon, 17 Jun 1996 09:19:41 -0400
Date: Mon, 17 Jun 96 08:16 CDT
X-Sender: waltzmn@popmail.skypoint.com
Message-Id: 
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: tc-list@scholar.cc.emory.edu
From: waltzmn@skypoint.com (Robert B. Waltz)
Subject: Re: Theories of trans.--continued
Sender: owner-tc-list@scholar.cc.emory.edu
Precedence: bulk
Reply-To: tc-list@scholar.cc.emory.edu
content-length: 3129

On Mon, 17 Jun 1996, DC PARKER  wrote:

>Dr Waltz writes that statistics are science, example is folklore.

First, let me apologize if I have given the impression that I have
a doctorate. I don't, and I don't have any formal training in
textual studies. I have a bachelor's in physics and math, and about
seven years' worth of fanatic interest in textual criticism. But
formal training -- none.

>Statistics is doubtless a kind of science, some examples are no
>doubt folklore.  But in our discipline (as in any other), statistics are
>meaningless without a skilled interpretation.  With regard to
>examples, what do we have _but_ examples?

Agreed in part; we have a finite number of manuscripts, with a finite
number of variations. So we cannot conduct "experiments" to gather
more data.

But the mass of data we already have is too much to be comprehended.
We *must* summarize. And I will maintain that the only way we can
summarize is statistically.

Let me give an example of the danger of examples. W. L. Richards, in
his extended study of the Johannine Epistles ("The Classification of
the Greek Manuscripts of the Johannine Epistles," SBL Dissertation
Series 35) claims that he can identify a manuscript's text-type
based on a mere five readings (pp. 188-189). But the strongly
Alexandrian ms. 81 has only three of these readings, and the purely
Alexandrian 33 has only *one* of them!

I agree that statistics need to be carefully prepared and intelligently
evaluated. (We all know how disastrous the "differences from TR"
statistic was for Streeter!) And examples are the only way to make
things clear. But *only* statistical methods can give reliable
results.

>I should also add that philology traditionally gave particular weight to
>the significant reading, one of which might prove a relationship
>between 2 mss more conclusively than forty less telling readings.
>Before anybody tells me that I'm referring to the study of the classic, I
>know.  But the principle has also been used in NT text. crit., and I
>think that we should keep it in mind.

Agreed -- but even the "significant reading" can be handled statistically.
The method I use is the "near-singular reading." In my tables of
results (such as the one I sent in my previous post), I give the
number of near-singular agreements between all witnesses. These are
generally *highly* significant readings. High rates of agreement here
are a very strong indicator of kinship. However, *low* rates of
agreement here do not necessarily imply lack of kinship. Like
most statistics, it is something we must use carefully.

>But we were not discussing the relationship between mss (in
>which statistics is used), but the evidence of our second century
>sources, which Bill had suggested had been overlooked in earlier
>mailings.

To date, we have not prepared statistics for most fathers. But
if good critical editions of the fathers were available, there
is no reason why we could not. With the fathers as with the
manuscripts, knowledge of the sources *must* precede final
decisions about readings.

Bob Waltz
waltzmn@skypoint.com

Back