Darwin-L Message Log 2:99 (October 1993)

Academic Discussion on the History and Theory of the Historical Sciences

This is one message from the Archives of Darwin-L (1993–1997), a professional discussion group on the history and theory of the historical sciences.

Note: Additional publications on evolution and the historical sciences by the Darwin-L list owner are available on SSRN.

<2:99>From WILLS@macc.wisc.edu  Mon Oct 18 22:14:03 1993

Date: Mon, 18 Oct 93 22:15 CDT
From: Jeffrey Wills <WILLS@macc.wisc.edu>
Subject: manuscripts, populations, horizontal transmission
To: darwin-l@ukanaix.cc.ukans.edu

Bob O'Hara wrote recently:

"There are many differences between manuscript transmission and both
linguistic and biological evolution, most notably that manuscript
transmission is not a populational phenomenon (it is closer to being
clonal, with a fair bit of horizontal transmission), but I will leave those
issues for a later discussion so as to emphasize the particular issue of

Perhaps "The time has come, the Walrus said . . ." to speak of population
phenomena and horizontal transmission.  Let me ask what "population phenomena"
and "horizontal transmission" mean?  I assume population refers to the fact
that species or languages on a tree represent group phenomena whereas
manuscript trees have discrete single entities as their "individuals" on the
tree. Although this is usually true, in large manuscript lineages we do refer
to clusters of similar manuscripts as families and occasionally treat the group
as a single entity. On those occasions however we usually start to feel there
is too much contamination to even draw a tree so we don't. As a result most of
the manuscript trees you will see are of closed traditions and tidy data sets.
This tends to bias the way we represent the manuscript history. Also, unlike
biologists who are interested in the whole tree, paleographers are usually
interested just in the best witnesses and later descendants are less studied
and often not represented on trees at all. This may blind us to issues we would
see if we studied the whole tree.

	The other term (horizontal transmission) is a familiar term in the world
of manuscripts and is used to describe the influence of a manuscript (s) other
than the parent manuscript on a copy. Does the term have a separate meaning in
	The suggestion that manuscript copying is clonal makes good sense: one
manuscript goes in, the same manuscript and another one come out. But as Bob
points out there is also coded material which is not from the source code and
is not random. How can something be part clonal and part not?
	This may seem bizarre, but I think we need to put the scribes into the
tree too. It's a terribly uneven sort of hybridization, but I think it has some
explanatory value. Example 1: I am copying along a passage in Virgil's Aeneid
which goes "de nomine Phoebi" (by the name of Apollo) and I write "de nomine
Christi" (by the name of Christ). This is not a mutation like transposing
letters or omitting a word, in which the scribe can be seen as a faulty
mechanic.  In the case of "de nomine Christi", the scribe has entered new code
into the tradition which is nowhere else attested in any manuscript of the
Aeneid.  Example 2: I am copying the copy generated above, which reads "de
nomine Christi". I know this is nonsense; the passage is talking about the
temple of Apollo; Virgil died before Christ, I think I remember the line, etc.
so I change this to "de nomine Phoebi".  Where did this "new" reading come
from? It was not in the text in front of me, and it is not random; in fact it
is ancestral. I must be carrying another copy of the text inside me.
	The reality is that texts are recorded in memory as well as in writing.
Potentially each scribe has a polyploid version of the text in memory.  Is it
impossible to consider each copying of a manuscript not as a cloning but as a
mating between the scribe's memory and the manuscript at hand? Although of
different natures apparently, both have a copy of something in the same code.
In most instances, the manuscript is so dominant and the scribe so recessive
that the issue of this mating looks almost entirely like the manuscript.  But
when I copy a manuscript some of that text is also copied into my memory. When
I next copy a manuscript of the same text, the physical manuscript may not be
so dominant and my internal version may not be so recessive. Or if I come
across a hole (lacuna) in the manuscript, my linguistic resources will be
	Traditionally, paleographers (scholars of manuscripts) see the tree as
constantly branching (a closed tradition) and are very embarrassed/confused by
any effort to connect nodes on the tree (called "contamination" or "horizontal
transmission").  Yet how does the code leap across from branch to branch if the
mutation process is strictly clonal?  I am very innocent of what "clonal" means
to a biologist, but I think it is the great mistake of our manuscript trees
(and our language trees, but that is a topic for another day) to assume they
are clonal. It boxes us into a process which is inherently faulty. It is true
that once we say manuscripts intermate with memories we have really grafted the
manuscript trees onto the language tree and that is a mess. But I think it is
what happens.
	In the past decade, for various literary projects thousands of megabytes
of text have been entered in machine readable-form. Experience has shown that
scanning is not good enough (because the cost of error-correction is too high)
and double manual entry (keyboard entry by two separate people whose results
are then matched for differences) offers the best results by price.  It is also
clear that the less the keyboarders know the language the better (we have
entered hundreds of megabytes of Greek texts through companies in the
Phillipines or Singapore). Accidental human error in copying always exists, but
those sources of error must be distinguished theoretically from "intelligent"
error.  Possibly there is no biological counterpart for this intelligent error,
but I think it is given some account by a view at mating rather than cloning.

Jeffrey Wills

Your Amazon purchases help support this website. Thank you!

© RJO 1995–2022