On Thu, 31 Oct 1996, Dave Washburn wrote:
> Hort developed his whole text-type system
> based on the principle "identity of error implies identity of
> origin," but I'm not sure that's a valid canon. How much "identity
> of error"? What is meant by "identity of origin?"
This is most definitely a valid canon when dealing with a stemmatic or
genealogical approach to the NT MSS. The amount of identity in error has
to be more than might occur by mere coincidence or accident (e.g. the very
few Armenian and Spanish Hebrew MS joint readings), and thus requires a
sufficient amount of data. "Identity of origin" is relative only to the
archetype of the MSS in question which possess the common errors.
The problem with Hort's use of this valid principle is that Hort virtually
turned the principle on its head and based his "genealogical" method on
the false (or "less-true" if one wants to be politically correct)
principle that "identity of _reading_ implies identity of origin". From
this, Hort moved his entire methodology into grouping MSS based upon
shared readings, regardless of whether such were genealogically-
significant "errors" or not (in fact, most of them were not).
By this shift, Hort was then able to reduce all MSS possessing a close
identity of reading into a single archetype, to be played off against
other single archetypes (i.e. neutral, Alexandrian, Western, Syrian [=
Byzantine]), and thus bypass all need for any _real_ genealogical research
from the first. His "evidence of groups" reflects precisely this shift
from "identity of error" as a basis for genealogical research into
"identity of reading". The excuse of course could be easily rationalized:
if the original text only is genuine, then all non-original readings are
"errors"; however, this is most certainly _not_ what is intended by the
original stemmatic principle.
> It seems to me that what we need most in order to answer this
> question are some clear examples of direct descent and/or copying so
> that we can - at least for one period of the stream of transmission -
> examine the forces that were in action, i.e. how much change,
> correction by other mss., and all the rest. Without such clear
> examples, I'm not totally convinced that we can answer this question
> (though I remain hopeful that somehow we can).
Since we do have some MSS where direct descent is known to exist, as well
as the various family relationships where clear descent from a lost
exemplar is known, it should be fairly easy to compare collation data and
see what factors played a part in the alteration of texts away from the
parent exemplar. Much of my own theory derives from what I have seen in
the errors, corrections, and alterations in MSS in general as compared
with what appears in the various family collation data.
Maurice A. Robinson, Ph.D. Professor of Greek and New Testament
Southeastern Baptist Theological Seminary Wake Forest, North Carolina